COGNITIVE THERAPY
OF ANXIETY DISORDERS
Cognitive Therapy
of Anxiety Disorders
Science and Practice
DAVID A. CLARK
AARON T. BECK
THE GUILFORD PRESS
New York
London
© 2010 The Guilford Press
A Division of Guilford Publications, Inc.
72 Spring Street, New York, NY 10012
www.guilford.com
All rights reserved
Except as indicated, no part of this book may be reproduced, translated, stored in a retrieval
system, or transmitted, in any form or by any means, electronic, mechanical, photocopying,
microfilming, recording, or otherwise, without written permission from the Publisher.
Printed in the United States of America
This book is printed on acid-free paper.
Last digit is print number:
9 8
7 6
5 4
3 2
1
These materials are intended for use only by qualified mental health professionals.
The Publisher grants to individual purchasers of this book nonassignable permission
to reproduce all materials for which photocopying permission is specifically granted
in a footnote. This license is limited to you, the individual purchaser, for personal
use or use with individual clients. This license does not grant the right to reproduce
these materials for resale, redistribution, electronic display, or any other purposes
(including but not limited to books, pamphlets, articles, video- or audiotapes, blogs,
file-sharing sites, Internet or intranet sites, and handouts or slides for lectures,
workshops, webinars, or therapy groups, whether or not a fee is charged). Permission
to reproduce these materials for these and any other purposes must be obtained in
writing from the Permissions Department of Guilford Publications.
Library of Congress Cataloging-in-Publication Data
Clark, David A., 1954Cognitive therapy of anxiety disorders: science and practice / David A. Clark
and Aaron T. Beck.
p. ; cm.
Includes bibliographical references and index.
ISBN 978-1-60623-434-1 (hardcover: alk. paper)
1. Anxiety disorders—Treatment. 2. Cognitive therapy. I. Beck, Aaron T. II. Title.
[DNLM: 1. Anxiety Disorders—therapy. 2. Cognitive Therapy—methods.
WM 172 C592c 2010]
RC531.C535 2010
616.85′22—dc22
2009027597
To my wife, Nancy, and our daughters,
Natascha and Christina, with sincere love
for your steadfast interest, support, and understanding
D. A. C.
To my wife, Phyllis,
our children, Roy, Judy, Daniel, and Alice,
and our grandchildren,
Jodi, Sarah, Andy, Debbie, Eric, Ben, Sam, and Becky,
with love
A. T. B.
About the Authors
David A. Clark, PhD, is Professor of Psychology at the University of New Brunswick,
Canada. He has published seven books, including Intrusive Thoughts in Clinical Disorders: Theory, Research, and Treatment; Cognitive-Behavioral Therapy for OCD; and
Scientific Foundations of Cognitive Theory and Therapy of Depression, as well as over
100 articles and chapters on various aspects of cognitive theory and therapy of depression and anxiety disorders. Dr. Clark is a Fellow of the Canadian Psychological Association, a Founding Fellow of the Academy of Cognitive Therapy, and a recipient of the
Academy’s Aaron T. Beck Award for significant and enduring contributions to cognitive
therapy. He is an Associate Editor of the International Journal of Cognitive Therapy
and maintains a private practice.
Aaron T. Beck, MD, is University Professor Emeritus of Psychiatry, School of Medicine,
University of Pennsylvania, and the founder of cognitive therapy. He has published 21
books and over 540 articles in professional and scientific journals. Dr. Beck is the recipient of numerous awards, including the Albert Lasker Clinical Medical Research Award
in 2006, the American Psychological Association Lifetime Achievement Award in 2007,
the American Psychiatric Association Distinguished Service Award in 2008, and the
Robert J. and Claire Pasarow Foundation Award for Research in Neuropsychiatry in
2008. He is President of The Beck Institute for Cognitive Therapy and Research and
Honorary President of the Academy of Cognitive Therapy.
vi
Preface
T
he intricacies of anxiety have continued to capture the attention of some of the world’s
greatest scientists, scholars, and critical thinkers. In 1953 Rollo May stated in Man’s
Search for Himself that the “middle of the twentieth century is more anxiety-ridden
than any period since the breakdown of the Middle Ages” (p. 30). If this statement characterized the last century, is it not even more applicable to the dawn of the 21st century
with all the social, political, and economic threats that besiege us? Despite an end to
the cold war, an era of relative global stability and cooperation, and an unprecedented
rise in economic prosperity and technological advances, many in the Western world live
in a state of perpetual threat and uncertainty. According to the National Institute of
Mental Health (2003) approximately 40 million American adults (18%) suffer from an
anxiety disorder, with serious mental illness, including the anxiety disorders, costing
an estimated $193 billion in lost personal earnings (Kessler et al., 2008). No wonder
the search for highly effective and accessible treatments for the anxiety disorders has
become a major health initiative for most developed countries.
Twenty-five years ago, coauthor Aaron T. Beck published Anxiety Disorders and
Phobias: A Cognitive Perspective with Gary Emery and Ruth Greenberg. In the first
part of that book, Beck introduced a cognitive model of anxiety disorders and phobias
that represented a significant reconceptualization of the etiology, nature, and treatment
of anxiety (Beck, Emery, & Greenberg, 1985). At that time, research on the cognitive
features of anxiety was scant, and so much of the theoretical scaffolding was, by necessity, based on clinical observation and experience. Since key aspects of the cognitive
model of anxiety had not yet been investigated, some of the treatment recommendations
described in the second half of the book have not stood the test of time. However, the
last 20 years has witnessed a virtual explosion in basic information-processing research
on the cognitive model of anxiety, the development of disorder-specific cognitive models and treatment protocols for the major anxiety disorders, and dozens of treatment
outcome studies demonstrating the efficacy of cognitive therapy of anxiety. In light of
the unprecedented advances in our understanding and treatment of the cognitive basis
of anxiety, a comprehensive, updated, and reformulated presentation of the cognitive
vii
viii
Preface
model of anxiety was needed so the model could be understood within the context of
contemporary research findings. This book, then, was born out of this necessity. In
addition, we believe that a single volume containing a detailed comprehensive treatment
handbook for cognitive therapy is timely in order to encourage greater use by clinicians
of evidence-based psychotherapy for the anxiety disorders.
The book is divided into three parts. Part I consists of four chapters on the reformulated cognitive model of anxiety and its empirical status. Chapter 1 discusses the distinctions between fear and anxiety and provides a rationale for taking a cognitive perspective on anxiety. Chapter 2 presents a reformulation of the generic cognitive model
of anxiety based on the original model (Beck et al., 1985) that was later refined by Beck
and Clark (1997). Twelve key hypotheses of the model are presented in Chapter 2, and
the vast empirical research relevant to these hypotheses is critically reviewed in Chapters 3 and 4. The literature review spans hundreds of studies conducted in key research
centers in Western Europe and North America, confirming our perception that the main
tenets of the cognitive model of anxiety have achieved a broad basis of empirical support.
The cognitive therapy approach has been applied to a wide range of psychiatric
and personality conditions. Thus, Part II consists of three chapters that explain how
the basic elements of cognitive therapy are used to alleviate anxiety. Chapter 5 reviews
several standardized measures of anxious symptoms and cognition that are useful for
assessment and treatment evaluation and provides a detailed explanation for producing a cognitive case formulation of anxiety. Chapters 6 and 7 present a step-by-step
description for implementing various cognitive and behavioral intervention strategies
for reduction of anxious symptoms. Case illustrations, suggested therapy narratives,
and clinical resource materials are provided in all three chapters as training tools in
cognitive therapy.
The final section, Part III, consists of five chapters that present disorder-specific
adaptations of cognitive therapy for panic disorder, social phobia, generalized anxiety
disorder, obsessive– compulsive disorder, and posttraumatic stress disorder. We excluded
specific phobias because there have been fewer developments on the cognitive aspects
of phobia since its presentation in Beck et al. (1985), and exposure-based treatment is
still considered the main treatment approach for reduction of phobic responses. Each of
the disorder-specific chapters presents a cognitive model tailored to that disorder and a
review of the empirical research that addresses key hypotheses of each model. In addition, the chapters offer disorder-specific case conceptualizations and cognitive therapy
strategies that target unique symptom features of each disorder. In essence, Part III
consists of five minitreatment manuals for complex anxiety disorders.
To assist therapists in explaining cognitive concepts and strategies to their clients,
we are in the process of developing a companion client workbook that will match the
organization and themes of the present book and will offer explanations for key aspects
of the therapy, homework exercises, and record-keeping forms.
We are indebted to a large contingent of renowned experts in the anxiety disorders
whose theoretical contributions, innovative and rigorous research, and clinically astute
treatment insights are responsible for the significant advances that we have presented in
this volume. In particular we acknowledge the notable contributions to cognitive theory and therapy of anxiety of Drs. Martin Antony, Jonathan Abramowitz, David Barlow, Thomas Borkovec, Brendan Bradley, Michelle Craske, David M. Clark, Meredith
Preface
ix
Coles, Michel Dugas, Edna Foa, Mark Freeston, Randy Frost, Richard Heimberg,
Stefan Hofmann, Robert Leahy, Colin MacLeod, Andrew Mathews, Richard McNally,
Karen Mogg, Christine Purdon, Stanley Rachman, Ronald Rapee, John Riskind, Paul
Salkovskis, Norman Schmidt, Robert Steer, Gail Steketee, Steven Taylor, and Adrian
Wells. Furthermore, we wish to acknowledge with gratitude the tenacity and meticulousness of Michelle Valley, who laboriously revised and validated all the references,
and to past and current graduate students, Mujgan Altin, Anna Campbell, Gemma
Garcia-Soriano, Brendan Guyitt, Nicola McHale, Adriana del Palacio Gonzalez, and
Adrienne Wang for their research and thoughtful discussions on cognitive aspects of
anxiety. We also appreciate the partial financial support for publication costs from the
University of New Brunswick Busteed Publication Fund. Finally we are grateful for the
encouragement, guidance, advice, and support of the staff at The Guilford Press, especially Jim Nageotte, Senior Editor, and Jane Keislar, Assistant Editor.
Contents
COGNITIVE THEORY AND R ESEARCH ON ANXIETY
1
Chapter 1.
Anxiety: A Common but Multifaceted Condition
3
Chapter 2.
The Cognitive Model of Anxiety
31
Chapter 3.
Empirical Status of the Cognitive Model of Anxiety
58
Chapter 4.
Vulnerability to Anxiety
PART I.
PART II.
101
COGNITIVE THERAPY OF ANXIETY:
ASSESSMENT AND INTERVENTION STRATEGIES
125
Chapter 5. Cognitive Assessment and Case Formulation
127
Chapter 6. Cognitive Interventions for Anxiety
180
Chapter 7.
Behavioral Interventions: A Cognitive Perspective
234
PART III.
COGNITIVE THEORY AND TREATMENT
OF SPECIFIC ANXIETY DISORDERS
273
Chapter 8. Cognitive Therapy of Panic Disorder
Chapter 9.
Cognitive Therapy of Social Phobia
275
332
Chapter 10. Cognitive Therapy of Generalized Anxiety Disorder
388
Chapter 11.
Cognitive Therapy of Obsessive–Compulsive Disorder
446
Chapter 12.
Cognitive Therapy of Posttraumatic Stress Disorder
491
References
557
Index
611
xi
PART I
COGNITIVE THEORY AND RESEARCH
ON ANXIETY
Cognitive therapy is a theory-driven psychotherapy with a strong commit-
ment to scientific empiricism. Its defining characteristics are not found in a set
of unique intervention strategies but rather in its cognitive conceptualization
of psychopathology and the therapeutic change process. Thus articulation of
the cognitive model as well as the derivation of testable hypotheses and their
empirical evaluation are critical to determining its construct validity. Similar
to the organization of earlier primary treatment manuals of cognitive therapy,
this book begins with a focus on the theoretical and empirical foundation of
cognitive therapy for anxiety. Chapter 1 discusses phenomenology, diagnostic
features, and the cognitive perspective on fear and anxiety. Chapter 2 presents
the reformulated generic or transdiagnostic cognitive model of anxiety and its
hypotheses, whereas Chapter 3 provides a critical evaluation of the prodigious
experimental literature relevant to key aspects of the cognitive model. This
section concludes with Chapter 4, which focuses on empirical evidence for cognitive vulnerability to experience heightened states of intense and persistent
anxiety.
1
Chapter 1
Anxiety
A Common but Multifaceted Condition
Love looks forward, hate looks back, anxiety has eyes
all over its head.
—M IGNON MCL AUGHLIN (American journalist, 1915– )
A
nxiety is ubiquitous to the human condition. From the beginning of recorded history, philosophers, religious leaders, scholars, and more recently physicians as well as
social and medical scientists have attempted to unravel the mysteries of anxiety and
to develop interventions that would effectively deal with this pervasive and troubling
condition of humanity. Today, as never before, calamitous events brought about by
natural disasters or callous acts of crime, violence, or terrorism have created a social
climate of fear and anxiety in many countries around the world. Natural disasters like
earthquakes, hurricanes, tsunamis, and the like have a significant negative impact on
the mental health of affected populations in both developing and developed countries
with symptoms of anxiety and posttraumatic stress showing substantial increases in the
weeks immediately following the disaster (Norris, 2005).
Elevated levels of anxiety and other posttraumatic symptoms spike in the first few
weeks after acts of terrorism, war, or other large-scale acts of community violence. In
5–8 weeks after the September 11, 2001, terrorist attacks on the World Trade Center
towers in New York City, symptoms of posttraumatic stress disorder (PTSD) doubled
(Galea et al., 2002). An Internet-based survey (N = 2,729) found that 17% of individuals outside New York City reported PTSD symptoms 2 months after 9/11 (Silver, Holman, McIntosh, Poulin, & Gil-Rivas, 2002). The National Tragedy Study, a telephone
survey of 2,126 Americans, found that 5 months after the 9/11 terrorist attacks month,
30% of Americans reported difficulty sleeping, 27% felt nervous or tense, and 17%
indicated they worried a great deal about future terrorist attacks (Rasinski, Berktold,
Smith, & Albertson, 2002). The Gallup Youth Survey of American teenagers conducted
2½ years after 9/11 found that 39% of teens were either “very” or “somewhat” worried
that they or someone in their families will become a victim of terrorism (Lyons, 2004).
3
4
COGNITIVE THEORY AND RESEARCH ON ANXIETY
Although large-scale threats have their greatest impact on the psychological morbidity
of individuals directly affected by the disaster in the weeks immediately following the
traumatic event, their wider effects are evident months and years later in the heightened
concerns and worries of a significant proportion of the general population.
Fear, anxiety, and worry, however, are not the exclusive domain of disaster and
other life-threatening experiences. In the majority of cases anxiety develops within the
context of the fluctuating pressures, demands, and stresses of daily living. In fact anxiety disorders represent the single largest mental health problem in the United States
(Barlow, 2002), with more than 19 million American adults having an anxiety disorder
in any given year (National Institute of Mental Health, 2001). Approximately 12–19%
of primary care patients meet diagnostic criteria for an anxiety disorder (Ansseau et al.,
2004; Olfson et al., 1997). Moreover, antidepressants and mood stabilizers are the third
most prescribed pharmacotherapy class, having 2003 global sales of $19.5 billion (IMS,
2004). Thus millions of people worldwide mount a daily struggle against clinical anxiety and its symptoms. These disorders cause a significant economic, social and health
care burden for all countries, especially in developing countries that face frequent social
and political upheavals and high rates of natural disaster.
This chapter provides an overview of the diagnosis, clinical features, and theoretical perspectives on the anxiety disorders. We begin by examining definitional issues
and the distinction between fear and anxiety. The diagnosis of anxiety disorders is
then considered with particular attention to the problem of comorbidity, especially with
depression and substance abuse disorders. A brief review of the epidemiology, course,
and consequence of anxiety is presented, and contemporary biological and behavioral
explanations for anxiety are considered. The chapter concludes with arguments for the
validity of a cognitive perspective for understanding the anxiety disorders and their
treatment.
ANXIETY AND FEAR
The psychology of emotion is rich with diverse and opposing views on the nature and
function of human emotions. All emotion theorists who accept the existence of basic
emotions, however, count fear as one of them (Öhman & Wiens, 2004). As part of our
emotional nature, fear occurs as a healthy adaptive response to a perceived threat or
danger to one’s physical safety and security. It warns individuals of an imminent threat
and the need for defensive action (Beck & Greenberg, 1988; Craske, 2003). Yet fear can
be maladaptive when it occurs in a nonthreatening or neutral situation that is misinterpreted as representing a potential danger or threat. Thus two issues are fundamental to
any theory of anxiety: how to distinguish fear and anxiety, and how to determine what
is a normal versus an abnormal reaction.
Defining Fear and Anxiety
Many different words in the English language relate to the subjective experience of anxiety such as “dread,” “fright,” “panic,” “apprehension,” “nervous,” “worry,” “fear,”
“horror,” and “terror” (Barlow, 2002). This has led to considerable confusion and
inaccuracy in the common use of the term “anxious.” However, “fear” and “anxiety”
Anxiety: A Common but Multifaceted Condition
5
must be clearly distinguished in any theory of anxiety that hopes to offer guidance for
research and treatment of anxiety.
In his influential volume on the anxiety disorders, Barlow (2002) stated that “fear
is a primitive alarm in response to present danger, characterized by strong arousal
and action tendencies” (p. 104). Anxiety, on the other hand, was defined as “a futureoriented emotion, characterized by perceptions of uncontrollability and unpredictability
over potentially aversive events and a rapid shift in attention to the focus of potentially
dangerous events or one’s own affective response to these events” (p. 104).
Beck, Emery, and Greenberg (1985) offered a somewhat different perspective on
the differentiation of fear and anxiety. They defined fear as a cognitive process involving “the appraisal that there is actual or potential danger in a given situation” (1985,
p. 8, emphasis in original). Anxiety is an emotional response triggered by fear. Thus
fear “is the appraisal of danger; anxiety is the unpleasant feeling state evoked when fear
is stimulated” (Beck et al., 1985, p. 9). Barlow and Beck both consider fear a discrete,
fundamental construct whereas anxiety is a more general subjective response. Beck et
al. (1985) emphasize the cognitive nature of fear and Barlow (2002) focuses on the
more automatic neurobiological and behavioral features of the construct. On the basis
of these considerations, we offer the following definitions of fear and anxiety as a guide
for cognitive therapy.
Clinician Guideline 1.1
Fear is a primitive automatic neurophysiological state of alarm involving the cognitive
appraisal of imminent threat or danger to the safety and security of an individual.
Clinician Guideline 1.2
Anxiety is a complex cognitive, affective, physiological and behavioral response system (i.e.,
threat mode) that is activated when anticipated events or circumstances are deemed to be
highly aversive because they are perceived to be unpredictable, uncontrollable events that
could potentially threaten the vital interests of an individual.
A couple of observations can be derived from these definitions. Fear as the basic
automatic appraisal of danger is the core process in all the anxiety disorders. It is evident in the panic attacks and acute spikes of anxiousness that people report in specific
situations. Anxiety, on the other hand, describes a more enduring state of threat or
“anxious apprehension” that includes other cognitive factors in addition to fear such as
perceived aversiveness, uncontrollability, uncertainty, vulnerability (helplessness), and
inability to obtain desired outcomes (see Barlow, 2002). Both fear and anxiety involve a
future orientation so that “what if?” questions predominate (e.g., “What if I ‘bomb’ this
job interview?”, “What if my mind goes blank during the speech?”, “What if my heart
palpitations trigger a heart attack?”).
The distinction between fear and anxiety can be illustrated by Bill, who suffers from
obsessive– compulsive disorder (OCD) due to a fear of contamination and so engages in
compulsive washing. Bill is hypervigilant about the possibility of encountering “dan-
6
COGNITIVE THEORY AND RESEARCH ON ANXIETY
gerous” contaminants, and so he avoids many things that he perceives as possible contamination. He is in a continual state of high arousal and subjectively feels nervous and
apprehensive due to repetitive doubts of contamination (e.g., “What if I become contaminated?”). This cognitive–behavioral–physiological state, then, describes anxiety. If
Bill touches a dirty object (e.g., the doorknob in a public building) he quickly experiences fear, which is the perception of imminent danger (e.g., “I’ve touched this dirty
doorknob. A cancer patient may have recently touched it. I could contract cancer and
die.”). Thus we describe Bill’s immediate response to the doorknob as “fear,” but his
almost continuous negative affective state as “anxiety.” Anxiety, then, is of greater concern for those individuals who seek treatment for a heightened state of “nervousness” or
agitation that causes considerable distress and interference in daily living. Consequently
it is anxiety and its treatment that is the focus of the present volume.
Normal versus Abnormal
It would be difficult to find someone who hasn’t experienced fear or felt anxious about
an impending event. Fear has an adaptive function that is critical to the survival of
the human species by warning and preparing the organism for response against lifethreatening dangers and emergencies (Barlow, 2002; Beck et al., 1985). Moreover, fears
are very common in childhood, and mild symptoms of anxiety (e.g., occasional panic
attacks, worry, social anxiety) are frequently reported in adult populations (see Craske,
2003, for review). So, how are we to distinguish abnormal from normal fear? At what
point does anxiety become excessive, so maladaptive that clinical intervention is warranted?
We suggest five criteria that can be used to distinguish abnormal states of fear and
anxiety. It is not necessary that all these criteria be present in a particular case, but one
would expect many of these characteristics to be present in clinical anxiety states.
1. Dysfunctional cognition. A central tenet of the cognitive theory of anxiety is
that abnormal fear and anxiety derive from a false assumption involving an erroneous
danger appraisal of a situation that is not confirmed by direct observation (Beck et al.,
1985). The activation of dysfunctional beliefs (schemas) about threat and associated
cognitive-processing errors leads to marked and excessive fear that is inconsistent with
the objective reality of the situation.
For example, the sight of a loose Rotweiller charging toward you with teeth bared
and raised fur on a lonely country road would likely elicit the thought “I am in grave
danger of being attacked; I better get out of here fast.” The fear experienced in this
situation is perfectly normal, because it involves a reasonable deduction based on an
accurate observation of the situation. On the other hand, anxiety elicited by the sight of
a toy poodle dog held on a leash by its owner is abnormal: the threat mode is activated
(e.g., “I’m in danger”) even though direct observation indicates this is a “nonthreatening” situation. In this latter case we would suspect that the person has a specific animal
phobia.
2. Impaired functioning. Clinical anxiety will directly interfere with effective and
adaptive coping in the face of a perceived threat, and more generally in the person’s daily
social or occupational functioning. There are instances in which the activation of fear
results in a person freezing, feeling paralyzed in the face of danger (Beck et al., 1985).
Barlow (2002) notes that rape survivors often report physical paralysis at some point
Anxiety: A Common but Multifaceted Condition
7
during the attack. In other cases the fear and anxiety may lead to a counterproductive
response that actually increases risk of harm or danger. For example, a woman anxious
about driving after being involved in a rear-end collision would constantly check her
rear-view mirror and so pay less attention to the traffic in front of her, increasing the
chance that she would cause the very accident she feared.
It is also recognized that clinical fear and anxiety usually interfere in a person’s
ability to lead a productive and fulfilling life. Consequently, in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR; American Psychiatric Association
[APA], 2000), marked distress or “significant interference with the person’s normal
routine, occupational (or academic) functioning, or social activities or relationships”
(p. 449) is one of the core diagnostic criteria for most of the anxiety disorders.
3. Persistence. In clinical states anxiety persists much longer than would be expected
under normal conditions. Recall that anxiety prompts a future-oriented perspective that
involves the anticipation of threat or danger (Barlow, 2002). As a result, the person with
clinical anxiety can feel a heightened sense of subjective apprehension by just thinking
about an impending potential threat, regardless of whether it eventually materializes.
Thus it is not uncommon for anxiety-prone individuals to experience elevated anxiety
on a daily basis over many years.
4. False alarms. In anxiety disorders one often finds the occurrence of false alarms,
which Barlow (2002) defines as “marked fear or panic [that] occurs in the absence of
any life-threatening stimulus, learned or unlearned” (p. 220). A spontaneous or uncued
panic attack is one of the best examples of a “false alarm.” The presence of panic
attacks or intense fear in the absence of threat cues or very minimal threat provocation
would suggest a clinical state.
5. Stimulus hypersensitivity. Fear is a “stimulus-driven aversive response” (Öhman
& Wiens, 2004, p. 72) to an external or internal cue that is perceived as a potential
threat. However, in clinical states fear is elicited by a wider range of stimuli or situations
of relatively mild threat intensity that would be perceived as innocuous to the nonfearful
individual (Beck & Greenberg, 1988). For example, most people would be quite fearful
about approaching a Sydney funnelweb spider, which has the most lethal spider venom
in the world for humans. On the other hand, a spider phobic patient was referred to
our clinical practice who exhibited intense anxiety, even panic attacks, at the sight of a
spider web produced by the smallest, most harmless Canadian household spider. Clearly
the number of spider-related stimuli that elicits a fear response in the phobic individual
is far greater than the spider-related stimuli that would elicit fear in the nonphobic individual. In the same way individuals with an anxiety disorder would interpret a broader
range of situations as threatening compared to individuals without an anxiety disorder.
Clinician Guideline 1.3 presents five questions to determine if a person’s experience of
fear or anxiety is sufficiently exaggerated and pervasive to warrant further assessment,
diagnosis, and possible treatment.
Clinician Guideline 1.3
1. Is fear or anxiety based on a false assumption or faulty reasoning about the potential for
threat or danger in relevant situations?
2. Does the fear or anxiety actually interfere in the person’s ability to cope with aversive or
difficult circumstances?
8
COGNITIVE THEORY AND RESEARCH ON ANXIETY
3. Is the anxiety present over an extended period of time?
4. Does the individual experience false alarms or panic attacks?
5. Is fear or anxiety activated by a fairly wide range of situations involving relatively mild
threat potential?
ANXIETY AND THE PROBLEM OF COMORBIDITY
Over the last several decades clinical research on anxiety has recognized that the older
term “anxiety neurosis” had limited heuristic value. Most theories and research on
anxiety now recognize that there are a number of specific subtypes of anxiety that
cluster under the rubric “anxiety disorders.” Even though these more specific anxiety
disorders share some common features such as the activation of fear in order to detect
and avoid threat (Craske, 2003), there are important differences with implications for
treatment. Thus the present volume, like most contemporary perspectives, will focus
on specific anxiety disorders rather than treat clinical anxiety as a single homogenous
entity. Table 1.1 lists the core threat and cognitive appraisal associated with the five
DSM-IV-TR anxiety disorders discussed in this book (for similar summary, see Dozois
& Westra, 2004).
Psychiatric classification systems like DSM-IV assume that mental disorders like
anxiety consists of more specific disorder subtypes with diagnostic boundaries that
sharply demarcate one type of disorder from another. However, a large body of epidemiological, diagnostic, and symptom-based research has challenged this categorical
approach to psychiatric nosology, offering much stronger evidence for the dimensional
nature of psychiatric disorders like anxiety and depression (e.g., Melzer, Tom, Brugha,
Fryers, & Meltzer, 2002; Ruscio, Borkovec, & Ruscio, 2001; Ruscio, Ruscio, & Keane,
2002).
One of the strongest challenges to the categorical perspective is the evidence of
extensive symptom and disorder comorbidity in both anxiety and depression—that
is, the cross-sectional co-occurrence of one or more disorders in the same individual
(Clark, Beck, & Alford, 1999). Only 21% of respondents with a lifetime history of disorder had only one disorder in the National Comorbidity Survey (NCS; Kessler et al.,
1994), a National Institute of Mental Health (NIMH) epidemiological study of mental
disorders involving a randomized nationally representative sample of 8,098 Americans
who were administered the Structured Clinical Interview for DSM-III-R. Based on a
sample of 1,694 outpatients from the Philadelphia Center for Cognitive Therapy evaluated between January, 1986, and October, 1992, only 10.5% of those with a primary
mood disorder and 17.8% with panic disorder (with or without agoraphobia avoidance)
had a “pure diagnosis” without Axis I or II comorbidity (Somoza, Steer, Beck, & Clark,
1994). Clearly then, diagnostic comorbidity is the norm rather than the exception, with
prognostic comorbidity, in which one disorder predisposes an individual to the development of other disorders (Maser & Cloninger, 1990) also important to consider in the
pathogenesis of psychiatric conditions.
Numerous clinical states have reported a high rate of diagnostic comorbidity within
the anxiety disorders. For example, a large outpatient study (N = 1,127) found that
Anxiety: A Common but Multifaceted Condition
9
TABLE 1.1. Core Features of Five DSM-IV-TR Anxiety Disorders
Anxiety disorder
Threatening stimulus
Core appraisal
Panic disorder
(with or without
agoraphobia)
Physical, bodily sensations
Fear of dying (“heart attack”),
losing control (“going crazy”) or
consciousness (fainting), having
further panic attacks
Generalized anxiety
disorder (GAD)
Stressful life events or other
personal concerns
Fear of possible future adverse or
threatening life outcomes
Social phobia
Social, public situations
Fear of negative evaluation from others
(e.g., embarrassment, humiliation)
Obsessive–compulsive
disorder (OCD)
Unacceptable intrusive
thoughts, images, or impulses
Fear of losing mental or behavioral
control or otherwise being responsible
for a negative outcome to self or others
Posttraumatic stress
disorder (PTSD)
Memories, sensations, external
stimuli associated with past
traumatic experiences
Fear of thoughts, memories,
symptoms, or stimuli associated with
the traumatic event
two-thirds of anxiety disorder patients had another current Axis I disorder, and over
three-fourths had a lifetime comorbid diagnosis (Brown, Campbell, Lehman, Grisham,
& Mancill, 2001). Individuals with an anxiety disorder, then, are much more likely
to have at least one or more additional disorders than would be expected by chance
(Brown et al., 2001).
Comorbid Depression
Anxiety disorders are more likely to co-occur with some disorders than with others.
Much of the research on comorbidity has focused on the relationship between anxiety
and depression. Approximately 55% of patients with an anxiety or depressive disorder
will have at least one additional anxiety or depressive disorder, and this rate jumps to
76% when considering lifetime diagnoses (Brown & Barlow, 2002). In the Epidemiologic Catchment Area (ECA) study individuals with a major depression were 9 to 19
times more likely to have a coexisting anxiety disorder than individuals without major
depression (Regier, Burke, & Burke, 1990). Fifty-one percent of anxiety disorder cases
in NCS had major depressive disorder, and this increased to 58% for lifetime diagnoses
(Kessler et al., 1996). Moreover, anxiety disorders are more likely to precede depressive
disorders than the reverse, although the strength of this sequential association does vary
across specific anxiety disorders (Alloy, Kelly, Mineka, & Clements, 1990; Mineka,
Watson, & Clark, 1998; Schatzberg, Samson, Rothschild, Bond, & Regier, 1998).
Results from the ECA survey waves indicated that simple phobia, obsessive– compulsive
disorder (OCD), agoraphobia, and panic attacks were associated with increased risk for
major depression 12 months later (Goodwin, 2002).
Research into comorbidity has important clinical implications for the treatment of
all psychological disorders. Clinical depression comorbid with an anxiety disorder is
associated with a more persistent course of disturbance, greater symptom severity, and
greater functional impairment or disability (Hunt, Slade, & Andrews, 2004; Kessler &
Frank, 1997; Kessler et al., 1996; Olfson et al., 1997; Roy-Byrne et al., 2000). In addi-
10
COGNITIVE THEORY AND RESEARCH ON ANXIETY
tion, anxiety disorders with a comorbid depression show a poorer treatment response,
higher relapse and recurrence rates, and greater service utilization than cases of pure
anxiety (Mineka et al., 1998; Roy-Byrne et al., 2000; Tylee, 2000).
Comorbid Substance Use
Substance use disorders, especially use of alcohol, are another category of conditions
that are often seen in the anxiety disorders. In their review Kushner, Abrams, and Borchardt (2000) concluded that presence of an anxiety disorder (except simple phobia)
doubles to quadruples the risk of alcohol or drug dependence, with anxiety frequently
preceding the alcohol use disorder and contributing to its persistence, although alcohol
misuse can also lead to anxiety. Even at subthreshold diagnostic levels, individuals with
an anxiety condition are significantly more likely to use drugs and alcohol than nonclinical controls (Sbrana et al., 2005).
It is evident that a special relationship exists between alcohol use disorders and
anxiety. Compared with mood disorders, anxiety disorders more often precede substance use disorders (Merikangas et al., 1998), leading to the assumption that anxious
individuals must be “self-medicating” with alcohol. However, this “self-medicating”
assumption was not supported in a 7-year prospective study in which alcohol dependence was as likely to increase risk of developing a subsequent anxiety disorder as was
the reverse temporal relationship (Kushner, Sher, & Erickson, 1999). Kushner and colleagues concluded that anxiety and alcohol problems likely have reciprocal and interacting influences that will lead to an escalation of both anxiety and problem drinking
(Kushner, Sher, & Beitman, 1990; Kushner et al., 2000). The end result can be a “downward self-destructive spiral” leading to helplessness, depression, and increased risk for
suicide (Barlow, 2002).
Comorbidity within Anxiety Disorders
The presence of one anxiety disorder significantly increases the probability of having
one or more additional anxiety disorders. In fact, pure anxiety disorders are less frequent than comorbid anxiety. In their large clinical study, Brown, DiNardo, Lehmann,
and Campbell (2001) found that comorbidity for another anxiety disorder ranged from
27% for specific phobia to 62% for posttraumatic stress disorder (PTSD). Generalized
anxiety disorder (GAD) was the most common secondary anxiety disorder, followed
by social phobia. For PTSD, which had the highest comorbid rate for another anxiety
disorder, panic disorder and GAD were the most common secondary anxiety conditions. Social phobia and GAD tended to precede many of the other anxiety disorders.
Analysis of lifetime diagnoses revealed even higher rates for occurrence of a secondary
anxiety disorder.
Clinician Guideline 1.4
A case conceptualization of anxiety should include a broad diagnostic assessment that covers investigation of comorbid conditions, especially major depression, alcohol abuse, and
other anxiety disorders.
Anxiety: A Common but Multifaceted Condition
11
PREVALENCE, COURSE, AND OUTCOME OF ANXIETY
Prevalence
The anxiety disorders are the most prevalent form of psychological disturbance (Kessler,
Chiu, Demler, & Walters, 2005). Epidemiological studies of adult community samples
have been remarkably consistent in documenting a 25–30% lifetime prevalence rate for
at least one anxiety disorder. For example the 1-year prevalence for any anxiety disorder
in the NCS was 17.2%, compared with 11.3% for any substance abuse/dependence and
11.3% for any mood disorder (Kessler et al., 1994). The NCS lifetime prevalence, which
includes all individuals who ever experienced an anxiety disorder, was 24.9%, but this
may be an underestimate because OCD was not assessed. In a recent replication of the
NCS (NCS-R), involving a nationally representative sample of respondents (N = 9,282)
interviewed between 2001 and 2003, 12-month prevalence for any anxiety disorder
was 18.1% and estimated lifetime prevalence was 28.8%, findings that are remarkably
similar to the first NCS (Kessler et al., 2005; Kessler, Berglund, Demler, Robertson, &
Walters, 2005).
National surveys conducted in other Western countries like Australia, Great Britain,
and Canada have also reported high rates of anxiety disorders in the general population,
although the actual prevalence rates vary slightly across studies because of different
interview methodologies, diagnostic decision rules, and other design factors (Andrews,
Henderson, & Hall, 2001; Jenkins et al., 1997; Canadian Community Health Survey,
2003). The World Health Organization (WHO) World Mental Health Survey Initiative
found that anxiety was the most common disorder in every country except the Ukraine
(7.1%), with 1-year prevalence ranging from 2.4% in Shanghai, China, to 18.2% in the
United States (WHO World Mental Health Survey Consortium, 2004).
Anxiety disorders are also common in childhood and adolescence, with 6-month
prevalence rates ranging from 6% to 17% (Breton et al., 1999; Romano, Tremblay,
Vitaro, Zoccolillo, & Pagani, 2001). The most frequent disorders are specific phobia,
GAD, and separation anxiety (Breton et al., 1999; Whitaker et al., 1990). Some disorders like social phobia, panic, and generalized anxiety significantly increase during
adolescence, whereas others like separation anxiety show a decrease (Costello, Mustillo,
Erkanli, Keeler, & Angold, 2003; Kashani & Orvaschel, 1990). Girls suffer higher rates
of anxiety disorders than boys (Breton et al., 1999; Costello et al., 2003; Romano et al.,
2001), comorbidity between anxiety and depression is high (Costello et al., 2003), and
anxiety disorders that arise during childhood and adolescence often persist into early
adulthood (Newman et al., 1996).
Individuals suffering from anxiety disorders often first come to the attention of
family physicians in primary care settings because of unexplained physical symptoms
like noncardiac chest pain, palpitations, faintness, irritable bowel syndrome, vertigo,
and dizziness. These complaints may reflect an anxiety condition such as panic disorder
(see discussion by Barlow, 2002). Moreover, patients with anxiety disorders seek out
medical advice in disproportionate numbers. Studies of primary care patients find that
10–20% have a diagnosable anxiety disorder (Ansseau et al., 2004; Olfson et al., 1997,
2000; Sartorius, Ustun, Lecrubier, & Wittchen, 1996; Vazquez-Barquero et al., 1997).
Sleath and Rubin (2002) found that anxiety was mentioned in 30% of visits to a university medical clinic family practice. Anxiety disorders, then, place a considerable burden
on health service resources.
12
COGNITIVE THEORY AND RESEARCH ON ANXIETY
A large percentage of the general adult population experiences occasional or mild symptoms of anxiety. There is some evidence that individuals are at increased risk for developing
a full-blown anxiety disorder if they experience panic attacks, sleep disturbance, or have
obsessional concerns that are not sufficiently frequent or intense to meet diagnostic criteria
(i.e., subclinical forms), or have high anxiety sensitivity (see Craske, 2003). Worry, the
cardinal feature of GAD, is reported by a majority of nonclinical individuals who express
concerns with work (or school), finances, family, and the like (e.g., Borkovec, Shadick, &
Hopkins, 1991; Dupuy, Beaudoin, Rhéaume, Ladouceur, & Dugas, 2001; Tallis, Eysenck,
& Mathews, 1992; Wells & Morrison, 1994). Problems with sleep are reported by 27%
of British women and 20% of British men (Jenkins et al., 1997). In the U.S. 1991 National
Sleep Foundation Survey, 36% of participants had occasional or chronic insomnia (AncoliIsrael & Roth, 1999). Other studies indicate that 11–33% of nonclinical students and community adults have experienced at least one panic attack in the last year (Malan, Norton,
& Cox, 1990; Salge, J. G. Beck, & Logan, 1988; Wilson et al., 1992). Thus symptoms of
anxiety and its disorders are prevalent problems that threaten the physical and emotional
well-being of a significant number of people in the general population.
Clinician Guideline 1.5
Given the high rate of anxiety disorders and symptoms in the general population, clinical
assessment should include specification of symptom frequency and intensity as well as measures that enable differential diagnosis between disorders.
Gender Differences
Women have a significantly higher incidence of most anxiety disorders than men (Craske,
2003), with the possible exception of OCD, where the rates are approximately equal (see
Clark, 2004). In the NCS women had a lifetime prevalence of 30.5% for any anxiety
disorder, compared with 19.5% for men (Kessler et al., 1994). Other community-based
and epidemiological studies generally have confirmed a 2:1 ratio of women to men in
prevalence of anxiety disorders (e.g., Andrews et al., 2001; Jenkins et al., 1997; Olfson
et al., 2000; Vazquez-Barquero et al., 1997). Since these gender differences were found
in community-based surveys, the preponderance of anxiety disorders in women cannot
be attributed to greater service utilization. In a critical review of research on gender
differences in the anxiety disorders, Craske (2003) concluded that women may have
higher rates of anxiety disorders because of an increased vulnerability such as (1) higher
negative affectivity; (2) differential socialization patterns in which girls are encouraged
to be more dependent, prosocial, empathic but less assertive and controlling of everyday
challenges; (3) more pervasive anxiousness as evidenced by less discriminating and more
overgeneralized anxious responding; (4) heightened sensitivity to reminders of threat
and contextual threat cues; and/or (5) tendency to engage in more avoidance, worry, and
rumination about potential threats.
Cultural Differences
Fear and anxiety exist in all cultures but their subjective experience is shaped by culturespecific factors (Barlow, 2002). Comparing the prevalence of anxiety across different cul-
Anxiety: A Common but Multifaceted Condition
13
tures is complicated by the fact that our standard diagnostic classification system, DSMIV-TR (APA, 2000), is based on American conceptualizations and experiences of anxiety
that may not have high diagnostic validity in other cultures (van Ommeren, 2002). Crosscultural generalizability is not necessarily improved by using the WHO’s classification of
anxiety disorders, the International Classification of Diseases—Tenth Revision (ICD-10),
because of the dominance of the European-influenced Western experience (World Health
Organization, 1992). Thus our standard diagnostic and assessment approaches to anxiety
may overemphasize aspects of anxiety that are prominent in the European Western experience and omit significant expressions of anxiety that are more culture-specific.
Barlow (2002) concluded in his review that apprehension, worry, fear, and somatic
arousal are common in all cultures. For example, a large community survey of 35,014
adult Iranians found that 20.8% had anxiety symptoms (Noorbala, Bagheri-Yazdi,
Yasamy, & Mohammad, 2004). Even in remote rural or mountainous regions of developing countries where modern industrial amenities and pressures are minimal, the occurrence of anxiety and panic disorders is similar to rates reported in Western community
surveys (Mumford, Nazir, Jilani, & Yar Baig, 1996). Nevertheless, countries do appear
to have different population rates of the anxiety disorders. The WHO World Mental
Health Surveys found that 1-year prevalence of DSM-IV anxiety disorders ranged from
a low of 2.4%, 3.2%, and 3.3% in Shanghai, Beijing, and Nigeria, respectively, to
11.2%, 12%, and 18.2% in Lebanon, France, and the United States, respectively (WHO
World Mental Health Survey Consortium, 2004). This broad variability in prevalence
rates raises the possibility that culture may influence the actual rate of anxiety disorders
across countries, although methodological differences across sites cannot be ruled out
as an alternative explanation for the differences.
There is substantial evidence that culture does play a significant role in the expression of anxious symptoms. Barlow (2002) noted that somatic symptoms appear more
prominent in emotional disorders in most countries other than those of the Europeaninfluenced West. Table 1.2 presents a select number of culture-bound syndromes with a
significant anxiety component.
Clinician Guideline 1.6
Assessment for anxiety should include a consideration of the individual’s culture and social/
familial environment and their influence on the development and subjective experience of
anxiety.
Persistence and Course
In contrast to major depression, anxiety disorders are often chronic over many years
with relatively low remission but more variable rates of relapse after complete recovery
(Barlow, 2002). The Harvard–Brown Anxiety Disorder Research Program (HARP), an
8-year prospective study, found that only one-third to one-half of patients with social
phobia, GAD, or panic disorder achieved full remission (Yonkers, Bruce, Dyck, & Keller,
2003).1 The Zurich Cohort Study found that nearly 50% of individuals with an initial
1 Although
these remission rates are very low, especially for social phobia and panic disorder, they probably
overestimate the true remission rates for the anxiety disorders since 80% of the subjects had some form of
pharmacological treatment over the 8-year follow-up.
14
COGNITIVE THEORY AND RESEARCH ON ANXIETY
TABLE 1.2. Select Culture-Bound Syndromes in Which Anxious Symptoms Play a Prominent Role
Syndrome name
Description
Country
dhat
Severe anxiety about the loss of semen through nocturnal
emissions, urination, or masturbation. (Sumathipala,
Siribaddana, & Bhugra, 2004)
Males in India, Sri
Lanka, China
koro
Sudden and intense fear that one’s sexual organs will retract
into the abdomen eventually causing death. (APA, 2000)
Mainly occurs in males
in south and east Asia
pa-leng
Morbid fear of the cold and wind in which the individual
worries about further loss of body heat that could eventually
lead to death. The person wears several layers of clothes even
on warm days to keep out wind and cold. (Barlow, 2002)
Chinese cultures
taijin kyofusho
An intense fear that one’s body parts or functions are
displeasing, offensive, or embarrassing to other people by
their appearance, odor, facial expressions, or movements.
(APA, 2000).
Japan
anxiety disorder later developed depression alone or depression comorbid with anxiety
at a 15-year follow-up (Merikangas et al., 2003). A Dutch longitudinal study of 3,107
older individuals found that 23% of subjects with an initial DSM-III anxiety disorder
continued to meet criteria 6 years later, whereas another 47% suffered from subclinical
anxiety (Schuurmans et al., 2005). It is evident the anxiety disorders persist for many
years when not treated (Craske, 2003). Given that the majority of these disorders have
their onset in childhood and adolescence (Newman et al., 1996), the chronic nature of
anxiety is a significant component of its overall disease burden.
Clinician Guideline 1.7
Consider the chronicity of anxiety and its influence on the development of other conditions
when conducting a cognitive assessment. We can expect that early onset and a more persistent course would be more challenging for treatment.
Consequences and Outcome
The presence of an anxiety disorder, or even just anxious symptoms, is associated with
a significant reduction in quality of life as well as in social and occupational functioning (Mendlowicz & Stein, 2000). In a meta-analytic review of 23 studies, Olatunji,
Cisler, and Tolin (2007) found that all individuals with anxiety disorders experienced
significantly poorer quality of life outcomes compared with control samples, and overall
quality of life impairment was equivalent across the anxiety disorders. Individuals with
an anxiety disorder have an increase in number of work loss days (Kessler & Frank,
1997; Olfson et al., 2000), more disability days (Andrews et al., 2001; Marcus, Olfson, Pincus, Shear, & Zarin, 1997; Weiller, Bisserbe, Maier, & LeCrubier, 1998), and
elevated rates of financial dependence in the form of disability payments, chronic unemployment, or welfare payments (Leon, Portera, & Weissman, 1995). Anxiety also tends
Anxiety: A Common but Multifaceted Condition
15
to reduce the quality of life and social functioning in patients with a comorbid chronic
medical illness (Sherbourne, Wells, Meredith, Jackson, & Camp, 1996). Olfson et al.
(1996) even found that primary care patients who did not meet diagnostic criteria for
GAD, panic, or OCD but had symptoms of these disorders reported significantly more
days of lost work, marital distress, and visits to a mental health professional. The negative impact of anxiety disorders in terms of distress, disability, and utilization of services
can be even greater than for individuals whose main problem is a personality disorder
or substance abuse (Andrews, Slade, & Issakidis, 2002). In fact, individuals with panic
disorder evidence significantly lower social and role functioning in daily activities than
patients with a chronic medical illness like hypertension (Sherbourne, Wells, & Judd,
1996).
Individuals with a diagnosable anxiety disorder make more visits to mental health
professionals and are more likely to consult with their general practitioners for psychological problems compared with nonclinical controls (Marciniak, Lage, Landbloom,
Dunayevich, & Bowman, 2004; Weiller et al., 1998). A large-scale study of employed
Americans found that individuals with anxiety disorders were significantly more likely
than the nonclinical control group to visit medical specialists, more likely to use inpatient services, and more likely to visit emergency rooms (Marciniak et al., 2004; see
also Leon et al., 1995, for similar results). However, the majority of individuals with an
anxiety disorder never receive professional treatment, and even fewer come to the attention of mental health practitioners (Coleman, Brod, Potter, Buesching, & Rowland,
2004; Kessler et al., 1994; Olfson et al., 2000). Family physicians, for example, are
particularly poor at recognizing anxiety, with at least 50% of anxiety disorders missed
in primary care patients (Wittchen & Boyer, 1998).
Given the adverse personal and social effects of anxiety disorders, the economic
costs of anxiety are substantial in both the direct costs of services and the indirect costs
of lost productivity. Self-reported anxiety in one American study accounted for an estimated 60.4 million days per year in lost productivity, which is equivalent to the level of
lost productivity associated with the common cold or pneumonia (Marcus et al., 1997).
Greenberg et al. (1999) estimated the annual cost of anxiety disorders at $42.3 billion
in 1990 U.S. dollars, whereas Rice and Miller (1998) found that the economic costs of
anxiety were greater than for schizophrenia or the affective disorders. 2
Clinician Guideline 1.8
Given the significant morbidity associated with anxiety, the negative impact of the disorder
on work/school productivity, social relations, personal finances, and daily functioning must
be included in the clinical assessment.
2 There
is evidence that a significant offset of the costs of anxiety can be achieved by early detection and
treatment (Salvador- Carulla, Segui, Fernández-Cano, & Canet, 1995). Health economic studies have consistently shown that cognitive-behavioral therapy (CBT) for anxiety disorders is cheaper than medication
and produces significant reduction in health care costs (Myhr & Payne, 2006). As the most common of the
mental disorders, anxiety inflicts a significant human and social cost on our society, but increased provision of cognitive and cognitive-behavioral treatment could reduce the personal and economic costs of these
disorders.
16
COGNITIVE THEORY AND RESEARCH ON ANXIETY
BIOLOGICAL ASPECTS OF ANXIETY
Anxiety is multifaceted, involving diverse elements of the physiological, cognitive,
behavioral, and affective domains of human function. Table 1.3 lists the symptoms of
anxiety divided into the four functional systems involved in an adaptive response to
threat and danger (Beck et al., 1985, 2005).
The automatic physiological responses that typically occur in the presence of threat
or danger are considered defensive responses. These responses, seen in the fear- eliciting
contexts of both animals and humans, involve autonomic arousal that prepares the
organism to deal with danger by fleeing (i.e., flight) or by directly confronting the danger
(i.e., fight), a process known as the “fight-or-flight” response (Canon, 1927). The behavioral features primarily involve escape or avoidance as well as safety-seeking responses.
The cognitive variables provide the meaningful interpretation of our internal state as
that of anxiety. Finally the affective domain is derived from cognitive and physiological
activation, and constitutes the subjective experience of feeling anxious. In the following sections, we briefly discuss the physiological, behavioral, and emotional aspects of
anxiety. The cognitive features of anxiety are the focus of subsequent chapters.
Psychophysiology
As evident from Table 1.3, many of the symptoms of anxiety are physiological in nature,
reflecting activation of the sympathetic (SNS) and parasympathetic (PNS) nervous systems. Activation of the SNS is the most prominent physiological response in anxiety,
and it leads to hyperarousal symptoms such as constriction of the peripheral blood
vessels, increased strength of the skeletal muscles, increased heart rate and force of contraction, dilation of the lungs to increase oxygen supply, dilation of the pupils for possible improved vision, cessation of digestive activity, increase in basal metabolism, and
increased secretion of epinephrine and norepinephrine from the adrenal medulla (BradTABLE 1.3. Common Features of Anxiety
Physiological symptoms
(1) Increase heart rate, palpitations; (2) shortness of breath, rapid breathing; (3) chest pain or pressure;
(4) choking sensation; (5) dizzy, lightheaded; (6) sweaty, hot flashes, chills; (7) nausea, upset stomach,
diarrhea; (8) trembling, shaking; (9) tingling or numbness in arms, legs; (10) weakness, unsteady,
faintness; (11) tense muscles, rigidity; (12) dry mouth
Cognitive symptoms
(1) fear of losing control, being unable to cope; (2) fear of physical injury or death; (3) fear of “going
crazy”; (4) fear of negative evaluation by others; (5) frightening thoughts, images, or memories; (6)
perceptions of unreality or detachment; (7) poor concentration, confusion, distractible; (8) narrowing of
attention, hypervigilance for threat; (9) poor memory; (10) difficulty in reasoning, loss of objectivity
Behavioral symptoms
(1) avoidance of threat cues or situations; (2) escape, flight; (3) pursuit of safety, reassurance; (4)
restlessness, agitation, pacing; (5) hyperventilation; (6) freezing, motionless; (7) difficult speaking
Affective symptoms
(1) nervous, tense, wound-up; (2) frightened, fearful, terrified; (3) edgy, jumpy, jittery; (4) impatient,
frustrated
Anxiety: A Common but Multifaceted Condition
17
ley, 2000). All of these peripheral physiological responses are associated with arousal
but cause various perceptible symptoms such as trembling, shaking, hot and cold spells,
heart palpitations, dry mouth, sweating, shortness of breath, chest pain or pressure, and
muscle tension (see Barlow, 2002).
The role of PNS excitation, which causes a conservation of certain physiological
responses, has not been as well researched in anxiety. The PNS is involved in symptoms like tonic immobility, drop in blood pressure, and fainting, which are a type of
“conservation-withdrawal” response strategy (Friedman & Thayer, 1998). The effects
of PNS stimulation include decreased heart rate and force of contraction, constricted
pupils, relaxed abdominal muscles, and constriction of the lungs (Bradley, 2000). Moreover, research on heart rate variability in panic attacks indicates that the cardiovascular
activity associated with anxiety should not be seen simply in terms of excessive SNS
activation but also reduced compensatory PNS excitation. Thus the PNS probably plays
a greater role in anxiety than previously considered.
Barlow (2002) concluded that one of the most robust and enduring findings in
the past 50 years of psychophysiological research is that chronically anxious individuals exhibit a persistently elevated autonomic arousal level often in the absence of an
anxiety-producing situation. For example, Cuthbert et al. (2003) reported significantly
elevated heart rate base levels for panic and specific phobias but not social phobia or
posttraumatic stress disorder (PTSD) groups. Other researchers, however, have linked
anxiety (or neuroticism) to excess autonomic lability and reactivity rather than to enduring tonic levels of activation (Costello, 1971; Eysenck, 1979). Craske (2003) proposed
that heightened cardiovascular reactivity might be a predisposing factor for panic disorder such that a tendency to experience intense and acute autonomic activation could
increase the salience and therefore threat attributed to bodily sensations.
Empirical support for autonomic differences between anxious and nonanxious controls in response to stressful or threatening stimuli has not been consistently obtained
across studies (Barlow, 2002). Freidman and Thayer (1998) also noted that psychophysiological findings of reduced heart rate and electrodermal variability challenge the view
that anxiety is characterized by excessive autonomic lability and reactivity. Nevertheless, anxious individuals do show a slower decline in their physiological response to
stressors (i.e., slow habituation), but this is probably due to their higher initial baseline
arousal level (Barlow, 2002). In addition Lang and colleagues found greater physiological arousal to fear-relevant imagery in snake phobic individuals, but reactivity was less
evident in those with panic (Cuthbert et al., 2003; Lang, 1979; Lang, Levin, Miller, &
Kozak, 1983). Together these results suggest that heightened physiological reactivity
to fear stimuli may be greatest in specific phobic conditions but less evident in other
anxiety states like panic disorder or PTSD. However, a heightened basal arousal level
and slower habituation rate might be seen more consistently across various anxiety disorders, thereby providing the physiological basis for chronically anxious individuals to
misinterpret their persistent state of hyperarousal as evidence of an anticipated threat
or danger.
Recent psychophysiological research suggests that individuals with chronic anxiety exhibit diminished autonomic flexibility in response to stressors (Noyes & HoehnSaric, 1998). This is characterized by a weak but sustained response to stressors, indicating a poor habituation trajectory. In a study of heart rate reactivity under baseline,
relaxation, and worry conditions, Thayer, Friedman, and Borkovec (1996) found that
18
COGNITIVE THEORY AND RESEARCH ON ANXIETY
individuals with GAD or those actively engaged in worry had lower cardiac vagal control, which supports the view that GAD is characterized by autonomic inflexibility.
In sum it would appear that important psychophysiological features of anxiety such
as elevated basal arousal level, slower habituation, and diminished autonomic flexibility
might contribute to the misinterpretation of threat that is the core cognitive feature of
anxiety. However, a different physiological response pattern may distinguish phobia,
panic disorder, and GAD, which prevents generalizing research findings across the anxiety disorders. Furthermore, it is unclear whether the anxiety state is primarily an excess
of SNS activation and a withdrawal of vagal activity, or if SNS activity is depressed and
PNS activity remains normal under the conditions of daily living (see Mussgay & Rüddel, 2004, for discussion).
Clinician Guideline 1.9
Assessment of anxiety disorders must include a thorough evaluation of the type, frequency,
and severity of physiological symptoms experienced during acute anxiety episodes, as well
as the patient’s interpretation of these symptoms. Baseline as well as patterns of physiological
reactivity should be assessed using diaries and daily rating scales.
Genetic Factors
There is considerable empirical evidence that anxiety runs in families (see Barlow, 2002,
for review). In a meta-analysis of family and twin studies for panic disorder, GAD,
phobias, and OCD, Hettema, Neale, and Kendler (2001) concluded there is significant
familial aggregation for all four disorders, with the strongest evidence for panic disorder. Across all disorders, estimates of heritability ranged from 30 to 40%, leaving the
largest proportion of the variance due to individual environmental factors. Even at the
symptomatic level, heritability accounts for only 27% of the variability by predisposing
individuals to general distress, with environmental factors determining the development
of specific anxiety or depressive symptoms (Kendler, Heath, Martin, & Eaves, 1987).
Barlow (2002) raised the possibility that a separate genetic transmission might be
evident for anxiety and panic. In a structural equation modeling of diagnostic data collected on a large female twin sample, Kendler et al. (1995) found separate genetic risk
factors for major depression and GAD (i.e., anxiety), on the one hand, and for acute,
short-lived anxiety like phobias and panic, on the other. An earlier study also found
a common genetic diathesis for major depression and GAD with disorder-specificity
determined by exposure to different life events (Kendler, Neale, Kessler, Heath, &
Eaves, 1992a).
There is less evidence that individuals inherit specific anxiety disorders and stronger
empirical support for inheritance of a general vulnerability to develop an anxiety disorder (Barlow, 2002). This nonspecific vulnerability for anxiety could be neuroticism,
high trait anxiety, negative affectivity, or what Barlow, Allen, and Choate (2004) called
a “negative affect syndrome.” Vulnerable individuals might show a stronger (or at least
more sustained) emotional response to aversive or stressful situations. However, environmental and cognitive factors would interact with this genetic predisposition to determine which of the specific anxiety disorders is experienced by a particular individual.
Anxiety: A Common but Multifaceted Condition
19
Clinician Guideline 1.10
A diagnostic interview should include questions about the prevalence of anxiety disorders
in first-degree relatives.
Neurophysiology
In the last decade rapid advances have been made in our understanding of the neurobiological basis of fear and anxiety. One important finding that has emerged is the central
role of the amygdala in emotional processing and memory (see discussion by Canli et al.,
2001). Human and nonhuman research indicates that the amygdala is involved in the emotional modulation of memory, the evaluation of stimuli with affective significance, and
the appraisal of social signals related to danger (see Anderson & Phelps, 2000). Research
on auditory fear conditioning by LeDoux (1989, 1996, 2000) has contributed most to
implicating the amygdala as the neural substrate for the acquisition of conditioned fear
responses. LeDoux (1996) concluded that the amygdala is the “hub in the wheel of fear”
(p. 170), that it is “in essence, involved in the appraisal of emotional meaning” (p. 169).
LeDoux (1989) contends that one of the most important tasks of the emotional brain
is to evaluate the affective significance (e.g., threat vs. nonthreat) of mental (thoughts,
memories), physical, or external stimuli. He proposed two parallel neural pathways in
the amygdala’s processing of fear stimuli. The first pathway involves direct transmission
of a conditioned fear stimulus through the sensory thalamus to the lateral nucleus of
the amygdala, bypassing the cortex. The second pathway involves transmission of fear
stimulus information from the sensory thalamus through the sensory cortex and on
to the lateral nucleus. Within the amygdala region the lateral nucleus, which receives
inputs in fear conditioning, innervates the central nucleus that is responsible for the
expression of the conditioned fear response (see also Davis, 1998). Figure 1.1 illustrates
the two parallel pathways of LeDoux’s conditioned fear reaction system.
LeDoux (1996) draws a number of implications from his dual pathway of fear. The
more direct thalamo– amygdala path (called “the low road”) is quicker, more rudimentary, and occurs without thinking, reasoning, and consciousness. The thalamo– cortical–
amygdala path (labeled “the high road”) is slower but involves more elaborative processing of the fear stimulus because of extensive involvement of higher cortical regions
of the brain. Although LeDoux (1996) discusses the obvious evolutionary advantage
of an automatic, preconscious neural basis to information processing of fear stimuli,
his research demonstrated that the cortical pathway is necessary for fear conditioning
to more complex stimuli (i.e., when the animal must discriminate between two similar
tones in which only one is paired with the unconditioned stimulus [UCS]).
The central role of the amygdala in fear is entirely consistent with its neuroanatomical connections. It has multiple output projections via the central nucleus to the
hypothalamus, hippocampus, and upward to various regions of the cortex, as well as
downward to various brainstem structures involved in autonomic arousal and neuroendocrine responses associated with stress and anxiety like the periaqueductal gray region
(PAG), the ventral tegmental area, the locus ceruleus, and the raphe nuclei (Barlow,
2002). All of these neutral structures have been implicated in the experience of anxiety,
including the bed nucleus of the stria terminalis (BNST; Davis, 1998), which may be the
most important neural substrate of anxiety (Grillon, 2002).
20
COGNITIVE THEORY AND RESEARCH ON ANXIETY
Cortico–amygdala pathway
(slow but more elaborated
processing)
Emotional
Stimulus
Sensory Cortex
Sensory
Thalamus
Thalamo–amygdala pathway
(rapid but crude processing)
Amygdala
Emotional
Response
FIGURE 1.1. LeDoux’s parallel neural pathways in auditory fear conditioning.
The role of conscious cognitive processing in fear is a much debated issue in light of
LeDoux’s research suggesting a rapid and rudimentary noncortical thalamo– amygdala
pathway in the processing of conditioned fear. In fact LeDoux (1996) found that fearrelevant stimuli can be implicitly processed by the amygdala through the subcortical
thalamo– amygdala pathway without conscious representation. Neuroimaging studies have found that fearful or negatively valenced stimuli are associated with relative
increases in regional cerebral blood flow (rCBF) in the secondary or associative visual
cortex and relative reductions in rCBF in the hippocampus, prefrontal, orbitofrontal, temporopolar, and posterior cingulated cortex (e.g., see Coplan & Lydiard, 1998;
Rauch, Savage, Alpert, Fishman, & Jenike, 1997; Simpson et al., 2000). These findings
have been interpreted as evidence that fear can be preconscious without the occurrence
of higher cognitive processing.
Evidence for a subcortical, lower order pathway to immediate conditioned fear processing should not divert attention away from the critical role that attention, reasoning,
memory, and subjective appraisal or judgments play in human fear and anxiety. LeDoux
(1996) found that the thalamo– cortico– amygdala pathway was activated in more complex fear conditioning. Moreover, the amygdala has extensive connections with the hippocampus and cortical regions, where it receives inputs from cortical sensory processing
areas, the transitional cortical area, and the medial prefrontal cortex (LeDoux, 1996,
2000). LeDoux emphasizes that the hippocampal system involving explicit memory and
the amygdala system involving emotional memory will be activated simultaneously by
the same stimuli and will function at the same time. Thus cortical brain structures
involved in working memory, such as the prefrontal cortex and the anterior cingulate
and orbital cortical regions, and structures involved in long-term declarative memory,
Anxiety: A Common but Multifaceted Condition
21
like the hippocampus and temporal lobe, are implicated with amygdala-dependent emotional arousal to provide the neural basis to the subjective (conscious) experience of
fear (LeDoux, 2000). The neural substrates of cognition, then, can be expected to play
a critical role in the type of fear acquisition and persistence that characterizes complex
human fears and anxiety disorders. This is supported by various neuroimaging studies
that found differential activation of various medial prefrontal and frontotempororbital
regions of the cortex (e.g., Connor & Davidson, 1998; Coplan & Lyiard, 1998; Lang,
Bradley, & Cuthbert; 1998; McNally, 2007; van den Heuvel et al., 2004; Whiteside,
Port, & Abramowitz, 2004).
In their review Luu, Tucker, and Derryberry (1998) argued that fear-relevant mental
representations of the cortex influence emotional functioning not only at the later stage
of fear expression and responsivity, but cortical influence can also serve an anticipatory
function even before sensory information is physically available. The authors conclude
that “with our highly evolved frontal networks, we humans are capable of cognitively
mediating our actions, and of inhibiting the more reflexive responses triggered by limbic and subcortical circuits” (Luu et al., 1998, p. 588). This sentiment was recently
echoed in a review paper by McNally (2007a) in which he concludes that activation in
the medial prefrontal cortex can suppress conditioned fear acquisition that is mediated
by the amygdala. Thus prefrontal executive functions (i.e., conscious cognitive processes) can have fear-inhibiting effects that involve learning new inhibitory associations
or “safety signals” that suppress fear expression (McNally, 2007a). Frewen, Dozois, and
Lanius (2008) concluded in their review of 11 neuroimaging studies of psychological
interventions for anxiety and depression that CBT alters functioning in brain regions
such as the dorsolateral, ventrolateral, and medial prefrontal cortices; anterior cingulate; posterior cingulate/precuneus; and the insular cortices that are associated with
problem solving, self-referential and relational processing, and regulation of negative
affect. Clearly, then, the extensive involvement of higher order cortical regions of the
brain in emotional experiences is consistent with our contention that cognition plays an
important role in the production of anxiety and that interventions like cognitive therapy
can effectively inhibit anxiety by engaging cortical regions responsible for higher order
reasoning and executive function.
Neurotransmitter Systems
Neurotransmitter systems such as the benzodiazepine–gamma-aminobutyric acid
(GABA), noradrenergic, and serotonergic, as well as the corticotropin-releasing hormonal pathway, are important to the biology of anxiety (Noyes & Hoehn-Saric, 1998).
The serotonergic neurotransmitter system has become of increasing interest in research
on anxiety and panic. Serotonin acts as a neurochemical break on behavior, with blockage of serotonin receptors in humans associated with anxiety (Noyes & Hoehn-Saric,
1998). Although low levels of serotonin have been implicated as a key contributor to
anxiety, direct neurophysiological evidence is mixed on whether abnormalities in serotonin can be found in anxiety disorders like GAD compared to controls (Sinha, Mohlman, & Gorman, 2004). The serotonergic system projects to diverse areas of the brain
that regulate anxiety like the amygdala, septo-hippocampal, and prefrontal cortical
regions and so may have a direct influence on anxiety or an indirect influence by alter-
22
COGNITIVE THEORY AND RESEARCH ON ANXIETY
ing the function of other neurotransmitters (Noyes & Hoehn-Saric, 1998; Sinha et al.,
2004).
A subgroup of the inhibitory transmitter GABA contains benzodiazepine receptors
that enhance the inhibitory effects of GABA when benzodiazepine molecules bind to
these receptor sites (Gardner, Tully, & Hedgecock, 1993). Evidence that generalized
anxiety may be due to a suppressed benzodiazepine-GABA system comes from the anxiolytic effects of benzodiazepine drugs (e.g., lorazepam [Ativan], alprazolam [Xanax]),
which appear to have their clinical effectiveness by enhancing benzodiazepine-GABA
inhibition (Barlow, 2002).
Corticotropin-releasing hormone (CRH) is a neurotransmitter that is primarily stored in the hypothalamic paraventricular nuclei (PVN). Stressful or threatening
stimuli can activate certain brain regions like the locus ceruleus, amygdala, hippocampus, and prefrontal cortex, which then releases CRH. CRH then stimulates secretion
of adrenocorticotropic hormone (ACTH) from the anterior pituitary gland and other
pituitary– adrenal activity that results in increased production and release of cortisol
(Barlow, 2002; Noyes & Hoehn-Saric, 1998). The CRH, then, not only mediates endocrine responses to stress but also other broad brain and behavioral responses that play
a role in the expression of stress, anxiety, and depression (Barlow, 2002). Overall, then,
abnormalities at the neurotransmitter level appear to have anxiogenic or anxiolytic
effects that play an important contributory role in heightened physiological states that
characterize fear and anxiety. However, the exact nature of these abnormalities is still
unknown. Table 1.4 provides a summary of the biological aspects of anxiety that might
underlie the cognitive features of these disorders discussed later in this volume.
Clinician Guideline 1.11
Discuss the neural basis of anxiety when educating the client about the cognitive model of
anxiety. The rationale for cognitive therapy should include a discussion of how the higher
order cortical centers of the brain involved in memory, reasoning, and judgment can “override” or inhibit subcortical emotional brain structures, thereby reducing the subjective
experience of anxiety.
TABLE 1.4. Biological Concomitants of Cognition in Anxiety
Biological factors
Cognitive sequelae
•฀ Elevated tonic autonomic activation
•฀ Increased salience of threat-related stimuli
•฀ Slower habituation rate
•฀ Sustained attention to threat
•฀ Diminished autonomic flexibility
•฀ Reduced ability to shift attention
•฀ Genetic predisposition for negative emotionality
•฀ Hypervalent schemas of threat and danger
•฀ Subcortical fear potentiation
•฀ Preconscious fear stimulus identification and
immediate physiological arousal
•฀ Extensive cortical afferent and efferent pathways
to subcortical emotion-relevant circuitry
•฀ Cognitive appraisal and memory influence fear
perception and modulates fear expression and
action
Anxiety: A Common but Multifaceted Condition
23
BEHAVIORAL THEORIES
Over several decades experimental psychologists grounded in learning theory have demonstrated that fear responses can be acquired through an associative learning process.
Theoretical and experimental work from this perspective has focused on the physiological and behavioral responses that characterize an anxious or fearful state. Early
learning theory focused on the acquisition of fears or phobic reactions through classical
conditioning.
Conditioning Theories
According to classical conditioning, a neutral stimulus, when repeatedly associated with
an aversive experience (unconditioned stimulus [UCS] that leads to the experience of
anxiety (unconditioned response [UCR]), becomes associated with the aversive experience, it acquires the capability to elicit a similar anxiety response (conditioned response
[CR]) (Edelmann, 1992). The emphasis in classical conditioning is that human fears are
acquired as a result of some neutral stimulus (e.g., visit to a dentist’s office) coming into
association with some previous anxiety-provoking experience (e.g., a highly painful and
terrifying experience at the dentist office when a child). Although numerous experimental studies over the past 80 years have demonstrated that fears can be acquired in the
laboratory by repeatedly pairing a neutral stimulus (e.g., tone) with an unconditioned
stimulus (e.g., mildly aversive electric shock), the model could not provide a credible
explanation for the remarkable persistence of human fears in the absence of repeated
UCS–CS pairings (Barlow, 2002).
Mowrer (1939, 1953, 1960) introduced a major revision to the conditioning theory
in order to better account for avoidance behavior and the persistence of human fears.
Referred to as “two-factor theory,” it became a widely accepted behavioral account of
the etiology and persistence of clinical fears and anxiety states throughout the 1960s
and early 1970s (e.g., Eysenck & Rachman, 1965). Although no longer considered a tenable theory of anxiety, the two-factor theory is important for two reasons. First, many
of the behavioral interventions that have proven so effective in the treatment of anxiety
disorders had their origins in the two-factor model. And second, our current cognitive
models of anxiety were in large part born out of the criticisms and inadequacies of the
two-factor theory.
Figure 1.2 provides an illustration of how the two-factor theory might be used to
explain Freud’s case study of Little Hans (Freud, 1909/1955). Little Hans was a 5-yearold Austrian boy who developed a fear that a horse would bite him, and so experienced
considerable anxiety whenever he ventured outside for fear of seeing a horse. The onset
of the “horse phobia” occurred after he witnessed a large “bus-horse” fall down and
violently kick its feet in an effort to get up. Little Hans then became frightened that
horses, particularly those pulling carts, would fall down and bite him. (Of course Freud
interpreted the real source of Little Han’s phobia as his repressed sexual affection for his
mother and hostility toward his father that became transposed [displaced] onto horses.)
In the two-factor model, the first stage of fear acquisition is based on classical
conditioning. Little Hans experiences a traumatic event: seeing a large horse fall to the
street and thrash about violently (UCS). This elicits a strong fear response (UCR), so
that the sight of horses (CS) through association with the UCS is now capable of elic-
24
COGNITIVE THEORY AND RESEARCH ON ANXIETY
Stage One—
One—Acquisition of Fear
UCS
UCR
Sight of horse
falling down and
thrashing violently
Intense
fear
CS
Horse
CR
Learned
fear
response
Stage Two—
Two—Persistence of Fear
Avoidance
Reinforcement
Phobia
Active avoidance of
horses
Fear reduction
is reinforcing
Fear
persists
FIGURE 1.2. A two-factor theory of fear acquisition explanation of Freud’s case study of Little
Hans.
iting a CR (fear response). However, fear persistence is explained at the second stage
because of extensive avoidance of the CS. In other words, Little Hans stays indoors
and so avoids the sight of horses (the CS). Because avoidance of horses ensures that
Little Hans will not experience fear or anxiety, the avoidance behavior is negatively
reinforced. Avoidance is maintained because fear reduction is a powerful secondary
reinforcer (Edelmann, 1992). Furthermore, because he stays indoors, Little Hans fails
to learn that horses do not regularly fall down (i.e., he does not experience repeated CSonly presentations that would lead to extinction).
By the late 1970s serious problems were raised with the two-factor model explanation for human phobias (Rachman, 1976, 1977; see also Davey, 1997; Eysenck, 1979).
First, classical conditioning assumes that any neutral stimulus can acquire fear- eliciting
properties if associated with a UCS. However, this assumption was not supported in
aversive conditioning experiments in which some stimuli (e.g., pictures of spiders and
snakes) produced a conditioned fear response much more easily than other stimuli (e.g.,
pictures of flowers or mushrooms; for review, see Öhman & Mineka, 2001). Second,
many individuals who develop clinical phobias can not recall a traumatic conditioning
event. Third, there is considerable experimental and clinical evidence of nonassociative
learning of fears through vicarious observation (i.e., witnessing someone else’s trauma)
or informational transmission (i.e., when threatening information about specific objects
or situations is conveyed to the individual). Fourth, people often experience traumatic
events without developing a conditioned fear response (Rachman, 1977). Again the
two-factor model requires considerable refinement to explain why only a minority of
individuals develop phobias in response to a traumatic experience (e.g., painful dental
work). And finally, the two-factor theory has difficulty explaining the epidemiology
of phobias (Rachman, 1977). For example, fear of snakes is much more common than
dental phobia, and yet many more people experience the pain of dental work than are
bitten by snakes.
Anxiety: A Common but Multifaceted Condition
25
Although various refinements were proposed, it became clear that the two-factor
theory of conditioning was unable to explain the development and persistence of human
fears and anxiety disorders. Many behavioral psychologists concluded that cognitive
constructs were needed to provide an adequate account of the development and maintenance of anxiety, even phobic states (e.g., Brewin, 1988; Davey, 1997). A variety of
cognitive concepts were proposed (e.g., expectancies, self- efficacy, attentional bias, or
threat-related schemas) as mediators between the occurrence of a fear- eliciting stimulus and the anxious response (see Edelmann, 1992). Not all behavioral psychologists,
though, embraced cognitive mediation as a causal mechanism in the development of
anxiety. An example of a more “noncognitive” perspective is the fear module proposed
by Öhman and Mineka (2001).
The Fear Module
Öhman and Mineka (2001) state that because fear evolved as a defense against predators and other threats to survival, it involves a fear module composed of behavioral,
psychophysiological, and verbal- cognitive components. A fear module is defined as “a
relatively independent behavioral, mental, and neural system that is specifically tailored
to help solve adaptive problems encountered by potentially life-threatening situations in
the ecology of our distant forefathers” (Öhman & Mineka, 2001, p. 484).
They discuss four characteristics of the fear module. First, it is selectively sensitized
to respond to stimuli that are evoluntionarily prepotent because they posed particular
threats to the survival of our ancestors. They reviewed a large experimental literature
that demonstrated selective association in human aversive conditioning in which individuals evidence better conditioning and greater resistance to extinction for phylogenetic stimuli (e.g., slides of snakes or spiders) than for ontogenetic materials (e.g., slides
of houses, flowers, or mushrooms). Öhman and Mineka (2001) concluded that (1) evolutionarily prepared fear-relevant stimuli have preferential access to the human fear
module and (2) selective association of these prepared stimuli is largely independent of
conscious cognition.
A second characteristic of the fear module is its automaticity. Öhman and Mineka
(2001) state that because the fear module evolved to deal with phylogenetic threats to
survival, it can be automatically activated without conscious awareness of the triggering
stimulus. Evidence for automatic preconscious activation of fear includes physiological
fear response (e.g., SCR) to fear stimuli that are not consciously recognized, continued
conditioned fear response to nonreportable stimuli, and the acquisition of a conditioned
fear response to fear-relevant stimuli that were not amenable to conscious awareness.
A third feature is encapsulation. The fear module is assumed to be “relatively impenetrable to other modules with which it lacks direct connections” (Öhman & Mineka,
2001, p. 485) and so will tend to run its course once activated with few possibilities
that other processes can stop it (Öhman & Wiens, 2004). Even though the fear module
is relatively impenetrable to conscious influences, Öhman and Mineka argue that the
fear module itself can have a profound influence by biasing and distorting conscious
cognition of the threat stimulus. In support of their contention of the independence of
the fear module from the influence of conscious cognition, Öhman and Weins (2004)
cite evidence that (1) masking of stimuli affects conscious appraisals but not condi-
26
COGNITIVE THEORY AND RESEARCH ON ANXIETY
tioned responses (SCRs), (2) instructions that alter explicit UCS–CS expectancies do
not affect conditioned response to biological fear-relevant stimuli, (3) individuals can
acquire conditioned fear responses to masked stimuli outside conscious awareness, and
(4) conditioned fear responses to masked stimuli can affect conscious cognition in the
form of expectancy judgments.
A final characteristic is its specific neural circuitry. Öhman and Mineka (2001) consider the amygdala the central neural structure involved in the control of fear and fear
learning and contend that fear activation (i.e., emotional learning) occurs via LeDoux’s
(1996) subcortical, noncognitive thalamo– amygdala pathway, whereas cognitive learning occurs via the hippocampus and higher cortical regions. The authors contend that
the amygdala has more afferent than efferent connections to the cortex and so has more
influence on the cortex than the reverse. Based on this view of the neural structure of
the fear module, they conclude that (1) nonconscious activation of the amygdala occurs
via a neural route that does not involve the cortex, (2) this neural circuitry is specific to
fear, and (3) any conscious cognitive processes associated with fear are a consequence of
the activated fear module (i.e., amygdala) and thus play no causal role in fear activation.
Thus biased appraisals and beliefs are a product of automatic fear activation and the
production of psychophysiological and reflexive defensive responses (Öhman & Weins,
2004). Exaggerated beliefs in danger may play a role in maintaining anxiety over time
but they are the consequence rather than the cause of fear.
Clinician Guideline 1.12
Given the substantial evidence concerning the importance of learning in the development of
anxiety, the clinician should explore with patients past anxiety-related learning experiences
(e.g., trauma, life events, exposure to threat-related information).
THE CASE FOR COGNITION
Öhman and Mineka’s (2001) perspective on fear and anxiety is at variance with the
cognitive perspective advocated by Beck and colleagues (Beck et al., 1985, 2005;
Beck & Clark, 1997; D. M. Clark, 1999). Although they acknowledge that cognitive phenomena should be targeted in treatment because they play a key role in the
longer term maintenance of anxiety, they still consider anxious thinking, beliefs, and
processing biases a consequence of fear activation. Öhman and Mineka (2001) do not
consider conscious cognition critical in the pathogenesis of fear itself, which is contrary
to the conceptualization of fear that we offered earlier in this chapter. This noncognitive view of fear is evident in other learning theorists like Bouton, Mineka, and Barlow
(2001), who argue that interoceptive conditioning in panic disorder occurs without
conscious awareness and is quite independent of declarative knowledge systems. Nevertheless, we consider cognitive appraisal a core element of fear and critical to understanding the etiology, persistence, and treatment of anxiety disorders. This view is
based on several arguments.
Anxiety: A Common but Multifaceted Condition
27
Existence of Preconscious Cognition
Critics of cognitive models tend to overemphasize conscious awareness when discussing cognition, arguing that the substantial experimental evidence of conditioned fear
responses without conscious awareness fails to support basic tenets of the cognitive perspective (e.g.,, Öhman & Mineka, 2001). However, there is equally robust experimental
research demonstrating preconscious, automatic cognitive and attentional processing of
fear stimuli (see MacLeod, 1999; Wells & Matthews, 1994; Williams, Watts, MacLeod,
& Mathews, 1997). Thus the cognitive perspective on anxiety is misrepresented when
cognition is characterized only in terms of conscious appraisal.
Cognitive Processes in Fear Acquisition (i.e., Conditioning)
Öhman and Mineka (2001) argue that cognitive processes are a consequence of fear
activation and so play little role in their acquisition. However, over the last three decades
many learning theorists have argued that cognitive concepts must be incorporated into
conditioning models to explain the persistence of fear responses. Davey (1997), for
example, reviews evidence that outcome expectancies as well as one’s cognitive representation of the UCS will influence the strength of the fear CR in response to a CS. In
other words, CRs increase or decrease in strength depending on how the person evaluates the meaning of the UCS or trauma (see also van den Hout & Merckelbach, 1991).
According to Davey (1997), then, cognitive appraisal is a key element in Pavlovian fear
conditioning.
It has long been recognized that outcome expectancies (i.e., expectations that in a
particular situation a certain response will lead to a given outcome) play a critical role
in aversive conditioning (e.g., Seligman & Johnston, 1973; de Jong & Merckelbach,
2000; see also experiments on covariation bias by de Jong, Merckelbach, & Arntz,
1995; McNally & Heatherton, 1993). In his influential review paper Rescorla (1988)
argued that modern learning theory views Pavlovian conditioning in terms of learning the relations among events (i.e., associations) that must be perceived and that are
complexly represented (i.e., memory) by the organism. For most behaviorally oriented
clinical researchers, then, the acquisition and elicitation of fear and anxiety states will
involve learning contingencies that recognize the influence and importance of various
cognitive mediators (for further discussion, see van den Hout & Merckelbach, 1991).
Conscious Cognitive Processes Can Alter Fear Responses
Öhman and Mineka (2001) contend that the fear module is impenetrable to conscious
cognitive control. However, this view is difficult to reconcile with empirical evidence
that cognitive or informational factors can lead to a reduction in fear (see discussion
by Brewin, 1988). Even with exposure-based interventions, which are directly derived
from conditioning theory, there is evidence that long-term habituation of fear responses
requires conscious directed attention and processing of the fear-relevant information
(Foa & Kozak, 1986). Brewin (1988) succinctly makes a case for the influence of cognition on fear responses, stating that “a theory that assigns a role to conscious thought
processes is necessary to explain how people can alternately frighten and reassure them-
28
COGNITIVE THEORY AND RESEARCH ON ANXIETY
selves by thinking different thoughts, test out a variety of different coping responses, set
goals and reward or punish themselves depending on the outcome, etc.” (p. 46).
The Amygdala Is Not Specific to Fear
A central argument of Öhman and Mineka (2001) is that a direct thalamus– amygdala
link in fear activation and emotional learning accounts for the automaticity of the fear
module and so is dissociable from declarative acquisition of information via the hippocampus. Thus activation of the amygdala begins a fear response which then leads to
more complex cognition and memory processes via projections to the hippocampus and
higher cortical brain regions (see also Morris, Öhman, & Dolan, 1998).
Although experimental research has been quite consistent in showing amygdaloid
activation in the processing of fearful stimuli, there is evidence that the amygdala may
also be involved in other emotional functions such as the appraisal of the social and
emotional significance of facial emotions (Adolphs, Tranel, & Damasio, 1998; Anderson & Phelps, 2000). Neuroimaging studies suggest greater activation occurs in the prefrontal cortex, amygdala, other midbrain structures, and the brainstem when processing
any generally negative, arousing emotional stimuli, which suggests that the amygdala
and other structures involved in emotional processing may not be specific to fear but
rather to the valence of emotional stimuli (e.g., Hare, Tottenham, Davidson, Glover, &
Casey, 2005; Simpson et al., 2000; see also amygdala activation when processing sad
film excerpts, Lévesque et al., 2003). In addition the amygdala is responsive to positively
valenced stimuli, although this response seems to be more variable and elaborative in
nature than the fixed, automatic response seen to fear expressions (Somerville, Kim,
Johnstone, Alexander, & Whalen, 2004; see also Canli et al., 2002). Thus there is
experimental evidence that the amygdala may not be the seat of anxiety specifically but
an important neural structure of emotion processing more generally (see also Gray &
McNaughton, 1996).
Other neuroimaging research suggests that the amygdala can be influenced by
cognitive processes mediated by higher cortical regions of the brain. McNally (2007a)
reviewed evidence that the medial prefrontal cortex can suppress conditioned fear
acquired via activation of the amygdala. For example, in one study perceptual processing of threatening pictorial scenes was associated with a strong bilateral amygdala
response that was attenuated by cognitive evaluation of the fear stimuli (Hariri, Mattay,
Tessitore, Fera, & Weinberger, 2003). Together these findings suggest that conscious
cognitive processes mediated by other cortical and subcortical regions of the brain
have an important influence on the amygdala and together provide an integrated neural
account of the experience of fear.
Role of Higher Order Cortical Regions in Fear
The critical issue for a cognitive perspective on anxiety is whether conscious cognitive processes play a sufficiently important role in the propagation and amelioration of
anxiety to warrant an emphasis at the cognitive level. As discussed previously, there is
considerable neurophysiological evidence that higher cortical regions of the brain are
involved in the type of human fear and anxiety responses that are the target of clinical
Anxiety: A Common but Multifaceted Condition
29
interventions. LeDoux (1996) has shown that the hippocampus and related areas of the
cortex involved in the formation and retrieval of memories are implicated in more complex contextual fear conditioning. It is this type of conditioning that is particularly relevant to the formation and persistence of anxiety disorders. Moreover, LeDoux (1996,
2000) notes that the subjective feeling associated with fear will involve connections
between the amygdala and the prefrontal cortex, anterior cingulate, and orbital cortical regions, as well as the hippocampus. From a clinical perspective, it is the subjective
experience of anxiety that brings individuals to the attention of clinicians, and it is the
elimination of this aversive subjective state that is the main criteria for judging treatment success. In sum, it is apparent that the neural circuitry of fear is consistent with a
prominent role for cognition in the pathogenesis of anxiety.
SUMMARY AND CONCLUSION
In many respects anxiety is a defining feature of contemporary society and the tenacity of its clinical manifestations represents one of the greatest challenges facing mental
health research and treatment. The pervasiveness, persistence, and deleterious impact of
anxiety disorders have been well documented in numerous epidemiological studies. In
this chapter, a number of issues in the psychology of anxiety disorders were identified.
One of the most basic confusions arises from the definition of anxiety and its relation
to fear. Taking a cognitive perspective, we defined fear as the automatic appraisal of
imminent threat or danger, whereas anxiety is the more enduring subjective response
to fear activation. The latter is a more complex cognitive, affective, physiological, and
behavioral response pattern that occurs when events or circumstances are interpreted as
representing highly aversive, uncertain, and uncontrollable threats to our vital interests.
Fear, then, is the basic cognitive process underlying all the anxiety disorders. However,
anxiety is the more enduring state associated with threat appraisals, and so the treatment of anxiety has become a major focus in mental health.
Another fundamental issue associated with anxiety is the differentiation between
normal and abnormal states. Although fear is necessary for survival because it is essential for preparing the organism for response to life-threatening dangers, fear is clearly
maladaptive when present in the anxiety disorders. Once again a cognitive perspective
can be helpful in identifying the boundaries between normal anxiety or fear, and their
clinical manifestations. Fear is maladaptive and more likely associated with an anxiety disorder when it involves an erroneous or exaggerated appraisal of danger, causes
impaired functioning, shows remarkable persistence, involves a false alarm, and/or creates hypersensitivity to a wide range of threat-related stimuli. The challenge for practitioners is to offer interventions that “dampen down” or normalize clinical anxiety so it
becomes less distressing and interfering in daily living. The elimination of all anxiety
is neither desirable nor possible, but its reduction to within the normal range of human
experience is the common goal of treatment regimens for anxiety disorders.
Anxiety states are multifaceted, involving all levels of human function. There is a
significant biological aspect to anxiety, with particular cortical and subcortical neural
structures playing a critical role in emotional experience. This strong neurophysiological element gives anxiety states a sense of urgency and potency that makes modification
30
COGNITIVE THEORY AND RESEARCH ON ANXIETY
difficult. At the same time anxiety is often acquired through the organism’s interaction
with the environment even though this learning process may occur outside awareness
and beyond rational consideration. And yet cognitive mediation such as expectancies,
interpretations, beliefs, and memories play a critical role in the development and persistence of anxiety. As a subjective experience, anxiety may feel like a storm that surges
and recedes throughout the day. Relief from this state of personal turmoil can be a
potent motivator even when it elicits response patterns, such as escape and avoidance,
that are ultimately counterproductive to the vital interests of the individual.
Despite its complexity, we have argued in this chapter that cognition plays a key role
in understanding both normal and abnormal states of anxiety. The essence of maladaptive anxiety is a faulty or exaggerated interpretation of threat to an anticipated situation
or circumstance that is perceived to have significance for the person’s vital resources.
In the last two decades substantial progress has been made in elucidating the cognitive
structures and processes of anxiety. Based on the cognitive model of anxiety first proposed by Beck et al. (1985), this book presents a more refined, elaborated, and extended
cognitive formulation that incorporates major advances made within cognitive- clinical
research of anxiety. A systematic evaluation of the empirical status of this reformulation
is presented along with theory-driven strategies for cognitive assessment and treatment.
In subsequent chapters disorder-specific cognitive theories, research, and treatment are
presented for the major forms of anxiety disorders: panic disorder, social phobia, GAD,
OCD, and PTSD. It is our contention that the cognitive perspective continues to hold
much promise for the advancement of our understanding of anxiety and the provision
of innovative treatment approaches.
Chapter 2
The Cognitive Model of Anxiety
I
n cognitive therapy for anxiety and depression patients are taught a very basic idiom:
“The way you think affects the way you feel.” This simple statement is the cornerstone
of cognitive theory and therapy of emotional disorders, and yet individuals often fail to
recognize how their thoughts affect their mood state. Given the experience of intense
and uncontrollable physiological arousal often present during acute anxiety, it is understandable why those who suffer with it may not recognize its cognitive basis. Notwithstanding this failure in recognition, cognition does play an important mediational function between the situation and affect, as indicated in this diagram:
Triggering Situation
→฀ Anxious Thought/Appraisal →฀ Anxious Feeling
Individuals usually assume that situations and not cognitions (i.e., appraisals) are
responsible for their anxiety. Take, for example, how you feel in the period before an
important exam. Anxiety will be high if you expect the exam to be difficult and you
doubt your level of preparation. On the other hand, if you expect the exam to be quite
easy or you are confident in your preparation, anxiety will be low. The same holds true
for public speaking. If you evaluate your audience as friendly and receptive to your
speech, your anxiety will be lower than if you evaluate the audience as critical, bored,
or rejecting of your talk. In each example it is not the situation (e.g., writing an exam,
giving a speech, or having a casual conversation) that determines the level of anxiety,
but rather how the situation is appraised or evaluated. It is the way we think that has a
powerful influence on whether we feel anxious or calm.
The cognitive perspective can help us understand some apparent contradictions in
anxiety disorders. How is it possible for a person to be so anxious over an irrational and
highly improbable threat (e.g., that I might suddenly stop breathing), and yet react with
ease and no apparent anxiety in the face of more realistic dangers (e.g., developing lung
cancer from a chronic nicotine addiction)? What accounts for the highly selective and
situationally specific nature of anxiety? Why is anxiety so persistent despite repeated
nonoccurrences of the anticipated danger?
31
32
COGNITIVE THEORY AND RESEARCH ON ANXIETY
IMMEDIATE FEAR RESPONSE
CUE, OR STIMULUS
Orienting
Mode
Cognitive
processing
biases and
errors
PRIMAL THREAT
MODE
ACTIVATION
Immediate
defensive,
inhibitory responses
SECONDARY
ELABORATIVE
REAPPRAISAL
ANXIOUS SYMPTOMS
ACTIVATING SITUATION,
Increased
autonomic
arousal
STATE OF ANXIOUSNESS
Threat-oriented
thoughts and
images
FIGURE 2.1. Cognitive model of anxiety.
In this chapter we examine the nature and persistence of anxiety. We present the
cognitive model of anxiety as an explanation for one of the most important and perplexing questions faced by mental health researchers and practitioners: Why does anxiety
persist despite the absence of danger and the obvious maladaptive effects of this highly
aversive emotional state? The chapter begins with an overview of the cognitive model
(Figure 2.1) followed by a discussion of its central tenets, a description of the model,
analysis of the cognitive basis of normal and abnormal anxiety, and a statement of key
cognitive hypotheses.
OVERVIEW OF THE COGNITIVE MODEL OF ANXIETY
Anxiety: A State of Heightened Vulnerability
The cognitive perspective on anxiety centers on the notion of vulnerability. Beck, Emery,
and Greenberg (1985) defined vulnerability “as a person’s perception of himself as subject to internal or external dangers over which his control is lacking or is insufficient to
afford him a sense of safety. In clinical syndromes, the sense of vulnerability is magnified by certain dysfunctional cognitive processes” (pp. 67–68).
In anxiety this heightened sense of vulnerability is evident in individuals’ biased and
exaggerated appraisals of possible personal harm in response to cues that are neutral
or innocuous. This primary appraisal of threat involves an erroneous perspective in
which the probability that harm will occur and the perceived severity of the harm are
The Cognitive Model of Anxiety
33
greatly overestimated. Rachman (2004) noted that fearful individuals are much more
likely to overestimate the intensity of threat, which then leads to avoidance behavior. At
the same time anxious individuals fail to perceive the safety aspects of threat- evaluated
situations and tend to underestimate their ability to cope with the anticipated harm
or danger (Beck et al., 1985, 2005). This secondary elaborative reappraisal, however,
occurs immediately as a result of the primary threat appraisal, and in anxiety states it
amplifies the initial perception of threat. Thus the intensity of an anxiety state depends
on the balance between one’s initial appraisal of threat and the secondary appraisal of
coping ability and safety. The level or intensity of anxiety can be expressed in the following manner:
High Anxiety = ↑ threat probability/severity + ↓ coping and safety
Low Anxiety = ↓ threat probability/severity + ↑ coping and safety
Moderate Anxiety = ↔ threat probability/severity + ↔ coping and safety
Beck and Greenberg (1988) noted that the perception of danger sets off an “alarm
system” involving primal behavioral, physiological, and cognitive processes that evolved
to protect our species from physical harm and danger (see also Beck, 1985). Behavioral
mobilization to deal with the danger might involve a fight-or-flight response (escape
or avoidance), but it could also consist of other instrumental behaviors like calling for
assistance, taking a defensive stance, or negotiating to minimize the danger (Beck et
al., 1985, 2005). Autonomic arousal and other physiological responses that occur during threat vulnerability are important aspects of this early reflexive defense system.
The presence of anxiety activates behavioral mobilization to deal with perceived threat.
Although this primal behavioral mobilization evolved as a rapid and efficient response
to physical danger, it can impair actual performance when activated in benign situations
or the complex, diffusely stressful circumstances of contemporary society. Mobilization of the primal defense system can also have adverse effects when it is interpreted as
signaling a serious disorder such as when the person with panic disorder misinterprets
an elevated heart rate as a possible myocardial infarct (Beck et al., 1985; D. M. Clark
& Beck, 1988).
A second type of behavioral response often seen in anxiety states as a result of a
perception of threat is immobility in situations where active coping might increase the
actual or imagined danger (Beck et al., 1985). Signs of this immobility response may
be evident as freezing, feeling faint, or feeling “woozy.” It is associated with the cognitive perspective of being totally helpless. The immobility response is apparent in social
anxiety, such as when a highly anxious person feels faint when attempting to deliver a
public speech.
Despite the importance of behavioral mobilization and physiological arousal, it is
the initial primary appraisal of threat combined with a secondary appraisal of personal
inadequacy and diminished safety that are responsible for instigating anxiety. In this
sense faulty cognition is necessary but not sufficient for generating a state of anxiety.
The cognitive model of anxiety is rooted within an information-processing perspective, in which emotional disturbance occurs because of an excess or deficient functioning of the cognitive apparatus. Previously we defined information processing as “the
34
COGNITIVE THEORY AND RESEARCH ON ANXIETY
structures, processes, and products involved in the representation and transformation
of meaning based on sensory data derived from the external and internal environment”
(D. A. Clark et al., 1999, p. 77).
Anxiety, then, is the product of an information-processing system that interprets
a situation as threatening to the vital interests and well-being of the individual. In this
case a “threatening” meaning is generated and applied to the situation. The centrality of threat meaning-assignment (i.e., information processing) is nicely illustrated in
an example provided by Beck et al. (1985, 2005). Most individuals could easily walk
across a plank that is 6 inches wide without fear, if it were placed 1 foot off the ground.
However, raise the plank 100 feet off the ground, and most individuals would become
intensely afraid and refuse to walk the plank. What accounts for the different emotional
experiences in these two situations is that individuals evaluate walking a plank 100 feet
above the ground as highly dangerous. They also doubt whether their balance could be
maintained, and might actually experience dizziness and unsteadiness should they venture a few inches onto the plank. Although the plank is at different heights, their ability to elicit fear or anxiety depends on the perception of danger. Likewise perceptions
of danger are central to clinical states of anxiety. The cognitive model views clinical
anxiety as a reaction to an inappropriate and exaggerated evaluation of personal vulnerability derived from a faulty information-processing system that misconstrues neutral
situations or cues as threatening. This is entirely consistent with the definitions of fear
and anxiety proposed in Chapter 1. Based on the concept of vulnerability, Figure 2.1
illustrates the structures, processes, and products of the information-processing system
that are involved in the experience of anxiety.
Clinician Guideline 2.1
Correcting faulty appraisals of threat and secondary appraisals of vulnerability is a fundamental approach in cognitive therapy considered necessary for the reduction of anxiety.
Automatic and Strategic Processing
The cognitive model readily acknowledges that both automatic and strategic processes
are involved in anxiety (see Beck & Clark, 1997). Table 2.1 presents the defining characteristics of automatic and strategic or controlled processing first outlined in Beck and
Clark (1997).
At the cognitive level, automatic processing in anxiety has been most clearly demonstrated in the preconscious attentional bias for threat-related stimuli evidenced in
emotional Stroop and dot probe experiments (Macleod, 1999). Findings from implicit
memory tests suggest the presence of an automatic memory bias for negative information in anxiety disorders (Coles & Heimberg, 2002; Williams et al., 1997). Classical
conditioning research has demonstrated the acquisition of conditioned fear responses
(e.g., a skin conductance response) to masked fear-relevant stimuli presented outside
conscious awareness, indicating that fear learning can occur as an automatic, preconscious process (Öhman & Wiens, 2004). LeDoux’s (1996) research has documented the
acquisition of auditory fear responses in rodents via the subcortical thalamo– amygdala
pathway that bypasses the higher cortical centers for thinking, reasoning, and con-
The Cognitive Model of Anxiety
35
TABLE 2.1. Characteristics of Automatic and Strategic Processing
Automatic processing
Strategic (controlled) processing
•฀ Effortless
•฀ Effortful
•฀ Involuntary
•฀ Voluntary
•฀ Unintentional
•฀ Intentional
•฀ Primarily preconscious
•฀ Fully conscious
•฀ Fast, difficult to terminate or regulate
•฀ Slow, more amenable to regulation
•฀ Minimal attentional processing capacity
•฀ Requires a lot of attentional processing
•฀ Capable of parallel processing
•฀ Relies on serial processing
•฀ Stereotypic, involving familiar and highly
practiced tasks
•฀ Can deal with novel, difficult, and unpracticed
tasks
•฀ Low level of cognitive processing with minimal
analysis
•฀ Higher levels of cognitive processing involving
semantic analysis and synthesis
sciousness. Clearly, then, certain cognitive, neurophysiological, and learning processes
that are critical to the experience of anxiety occur at the automatic-processing level.
Although automatic processes are important to anxiety, one should not overlook
the central role played by the slower, more elaborative, and strategic processes in the
persistence of anxiety. Threat-biased judgments, reasoning, memory, and thinking are
critical parts of the subjective experience of anxiety that motivates individuals to seek
treatment. We should not overlook the importance of worry, anxious rumination, threat
images, and traumatic memories if we want to understand the anxiety disorders. In fact
controlled strategic processing allows us to interpret novel and complex information.
McNally (1995) concluded that, because of its meaning-assignment capabilities, strategic, elaborative processing is required for the anxious person to misinterpret innocuous
situations as threatening. Moreover, any particular cognitive task involves a mixture of
automatic and strategic processing, so a specific aspect of information processing should
not be rigidly dichotomized as automatic or strategic, but rather as reflecting more of
one type of processing than another (see McNally, 1995). Furthermore, involuntariness
rather than preconsciousness (i.e., outside conscious awareness) is the key feature of
automaticity in anxiety states (McNally, 1995; Wells & Matthews, 1994).
In the cognitive model (Figure 2.1) the initial orientation toward threat involves a
predominantly automatic, preconscious process. Activation of the primal threat mode
(i.e., the primary appraisal of threat) will be largely automatic because of the necessity
for rapid and efficient evaluation of a potential threat for the survival of the organism. (The term mode refers to a cluster of interrelated schemas organized to deal with
particular demands that pertain to one’s vital interests, survival, and adaptation [Beck,
1996; Beck et al., 1985, 2005; Clark et al., 1999].) However, some strategic, controlled
processing must occur even at this stage of the immediate threat response because of
our conscious, subjective experience of distress associated with the threat appraisal. As
we engage in secondary appraisal of coping resources, the presence or absence of safety,
and the reappraisal of the initial threat, this aspect of information processing will be
much more controlled, strategic, and elaborative. Even at this secondary stage responsible for a sustained anxiety response, processing will not be entirely strategic as evident
in processes such as worry and anxious rumination.
36
COGNITIVE THEORY AND RESEARCH ON ANXIETY
Clinician Guideline 2.2
Cognitive therapy teaches individuals to be more aware of their immediate threat appraisals
and to correct maladaptive secondary cognitive processes.
CENTRAL TENETS OF THE COGNITIVE MODEL OF ANXIETY
A number of propositions derived from the cognitive perspective guided the development of the cognitive model (see Figure 2.1). These propositions were first articulated in
the original cognitive model of anxiety (Beck et al., 1985, 2005) and are elaborated in
the sections below (see Table 2.2 for a definition of the basic tenets).
TABLE 2.2. Central Tenets of the Cognitive Model of Anxiety
Exaggerated threat appraisals
Anxiety is characterized by an enhanced and highly selective attention to personal risk,
threat, or danger that is perceived as having a serious negative impact on vital interests
and well-being.
Heightened helplessness
Anxiety involves an inaccurate evaluation of personal coping resources, resulting in an
underestimation of one’s ability to cope with a perceived threat.
Inhibitory processing of safety information
Anxiety states are characterized by inhibited or highly restricted processing of safety
cues and information that convey diminished likelihood and severity of a perceived
threat or danger.
Impaired constructive or reflective thinking
During anxiety more constructive, logical, and realistic elaborative thinking and
reasoning are difficult to access and so are ineffectively utilized for anxiety reduction.
Automatic and strategic processing
Anxiety involves a mixture of automatic and strategic cognitive processes that are
responsible for the involuntary and uncontrollable quality of anxiety.
Self-perpetuating process
Anxiety involves a vicious cycle in which heightened self-focused attention on the signs
and symptoms of anxiety will itself contribute to an intensification of subjective distress.
Cognitive primacy
The primary cognitive appraisal of threat and the secondary appraisal of personal
vulnerability can generalize such that a broader array of situations or stimuli are
misperceived as threatening and various physiological and behavioral defensive
responses are inappropriately mobilized to deal with the threat.
Cognitive vulnerability to anxiety
Increased susceptibility to anxiety is a result of enduring core beliefs (schemas) about
personal vulnerability or helplessness and the salience of threat.
The Cognitive Model of Anxiety
37
Exaggerated Threat Appraisals
We previously introduced the concept of exaggerated threat appraisal as a primary, core
feature of anxiety. The process of appraising or evaluating external or internal cues as
potential threat, danger, or harm to personal vital resources or well-being involves a
rapid, automatic, and highly efficient cognitive, physiological, behavioral, and affective
defensive system that evolved to protect and ensure the survival of the organism. Many
writers have noted the obvious evolutionary significance of a cognitive system primed
to rapidly and selectively scan the environment for anything that might pose a physical
danger to our primordial ancestors (Beck, 1985; D. M. Clark & Beck, 1988; Craske,
2003; Öhman & Mineka, 2001). Threat is rapidly appraised in terms of its temporal/
physical proximity or intensifying nature (i.e., “threat imminence” [Craske, 2003] or
“looming vulnerability” [Riskind & Williams, 2006]), probability of occurrence, and
severity of outcome. Together these evaluated characteristics of the stimulus will result
in the initial assignment of a threat value.
This primary assignment of threat value is inherent in all experiences of anxiety. In
the cognitive model this initial, relatively automatic threat appraisal is due to activation
of the primal threat mode (see Figure 2.1). The appraisal of threat will involve various
cognitive processes and structures including attention, memory, judgment, reasoning,
and conscious thought. This is illustrated in the following example. Imagine an individual running along a fairly isolated country road. He suddenly hears the bark of a dog
in the yard of a house he is approaching. Instantly his muscles tighten, his pace quickens,
his breathing and heart rate accelerate. These responses to the barking dog are triggered
by a very rapid initial threat appraisal that just barely registers in the runner’s conscious
awareness: “Am I in danger of an attack?” The situation will be assigned a high threat
value if the runner is close to the house in question, thinks there is high probability that
the dog is not leashed, and assumes the dog is large and vicious (high severity). On the
other hand, the runner might assign a low threat value with increased distance from the
dog, or if he concludes that the dog is probably leashed or simply a friendly household
pet. An immediate threat appraisal, then, will be apparent in all experiences of both
normal and abnormal anxiety states. In clinical anxiety, the primary threat appraisal is
exaggerated and disproportionate to the actual threat value of an event.
Clinician Guideline 2.3
Cognitive therapy focuses on helping clients recalibrate exaggerated threat appraisals and
increase their tolerance for risk and uncertainty related to their anxious concerns.
Heightened Helplessness
A secondary appraisal of personal resources and coping ability involves a more conscious, strategic evaluation of one’s ability to respond constructively to perceived threat.
This appraisal occurs at the secondary elaborative phase of the cognitive model (see
Figure 2.1). This secondary appraisal will involve Bandura’s (1977, 1989) concepts of
self-efficacy (“Do I have the ability to deal with this threat”?) and outcome expectancy
(“What is the likelihood that my efforts will reduce or eliminate the threat?”). Positive
38
COGNITIVE THEORY AND RESEARCH ON ANXIETY
self-efficacy and outcome expectation could lead to a reduction in anxiety, especially
if the person’s initial efforts to deal with the threat appeared successful. On the other
hand, low perceived self- efficacy and a negative outcome expectation would lead to a
heightened state of helplessness and greater feelings of anxiety.
Although secondary appraisal of coping resources is triggered by the primary threat
appraisal, both will occur almost simultaneously as a highly reciprocal and interactive
cognitive evaluation (Beck et al., 1985, 2005). As noted previously, the intensity of
anxiety will depend on the degree of threat in relation to one’s perceived capacity to
cope with the danger. In our case of the runner hearing a barking dog, anxiety would
be minimized if he recalled previous positive experiences of dealing with dogs, or if he
remembered he was carrying a can of pepper spray. In clinical anxiety individuals have a
heightened sense of helplessness in the face of certain perceived threats and so conclude
they are unable to deal with the anticipated danger.
Clinician Guideline 2.4
Increasing self-confidence to deal with threat and uncertainty is an important objective of
cognitive therapy of anxiety.
Inhibitory Processing of Safety
Beck (1985) noted that anxiety is not only characterized by a selective enhanced processing of danger but also a selective suppression of information that is incongruent
with perceived danger. D. M. Clark and Beck (1988) included underestimated rescue
factors (what others can do to help) as a cognitive error that will contribute to an exaggerated evaluation of threat in anxiety. It is suggested that in anxiety disorders the
immediate and automatic formation of a threat appraisal based on activation of threat
schemas will so bias the information-processing system toward detecting and evaluating threat, that any information incongruent with threat schemas will be filtered out,
even ignored. As a result any corrective information, which could lead to a reduction
in the threat value assigned to the situation, is lost and the anxiety persists. So, in our
example, a runner intensely anxious about the barking dog may fail to notice a fence
around the property, thus reducing the chance that the dog will charge out onto the
road. This apparent inability to process the safety aspects of a situation is clearly seen in
the anxiety disorders such as the speech-anxious person who fails to process cues from a
receptive audience, or the test-anxious student who has successfully answered the most
difficult questions.
Another consequence of inhibited processing of safety cues is that the person may
seek inappropriate ways to secure safety or avoid danger. The person with agoraphobia
may only venture outside with certain family members because this appears to reduce the
chance of a panic attack, or the individual with contamination obsessions may develop
certain compulsive rituals to reduce anxiety and secure a sense of safety from the prospect of contamination. Salkovskis (1996b) noted that safety-seeking behavior and avoidance may contribute to the persistence of anxiety, because both prevent disconfirmation
that the perceived threat is benign or will not occur. Thus in health anxiety the person
The Cognitive Model of Anxiety
39
may spend many hours searching the Internet for information that would confirm that
a particular skin rash is benign and not a sign of melanoma. However, in this case the
safety-seeking behavior (i.e., reassurance seeking) may be particularly maladaptive and
a potent contributor to anxiety because the individual fails to find conclusive evidence
to disconfirm the threat attributed to the skin rash. Another form of disconfirmation
bias occurs when the person with panic disorder, for example, engages in controlled
breathing (safety-seeking behavior) whenever he feels tightness in the throat and fears
suffocation. In this case the controlled breathing prevents the person from learning that
the throat sensation will not lead to the catastrophic outcome of suffocation.
Clinician Guideline 2.5
Improved processing of safety cues that disconfirm perceived threats is an important element in cognitive therapy of anxiety disorders.
Impaired Constructive or Reflective Thinking
During anxious states constructive modes of thinking are less accessible. This means
that slower, more logical and effortful deductive reasoning involving a more complete
and balanced processing of a situation’s threat potential is more difficult to achieve.
This more constructive, reflective approach to threat is under conscious control and
so takes more time and effort because it involves not only a more complete evaluation
of the threat and safety features of a situation, but it also requires selection of instrumental behaviors for dealing with anxiety. Beck et al. (1985, 2005) noted that this
constructive mode of thinking may be an alternative anxiety- reduction system to the
anxiety-potentiating, automatic primal threat process. However, this reasoned, elaborative cognitive orientation appears lost to individuals who are intensely anxious. The
predominance of the primal threat mode appears to inhibit access to constructive mode
thinking. Beck (1996) stated that once an automatic or primal mode of thinking is activated, it tends to dominant information processing until the activating circumstance
disappears.
The relative inaccessibility of constructive thinking contributes to the persistence of
anxiety. Beck (1987) argued that a key factor in the experience of panic is the inability
to realistically appraise (i.e., apply tests, draw on past experiences, generate alternative
explanations) a specific physical sensation (e.g., chest pain) in any way other than from
a catastrophic perspective. It is the existence of impaired reflective thinking that is a key
entry point for cognitive therapy of anxiety. Clients are taught cognitive restructuring
skills as a means of developing a more constructive cognitive perspective on perceived
threat.
Clinician Guideline 2.6
Cognitive therapy seeks to improve access to and the effectiveness of reflective thinking to
counter immediate faulty threat appraisals.
40
COGNITIVE THEORY AND RESEARCH ON ANXIETY
Automatic and Strategic Processing
We have already considered how automatic and strategic processes are evident at various facets of the cognitive basis of anxiety. Automatic processing will be more apparent
in the early primary appraisal of threat involving activation of the primal threat mode,
whereas controlled strategic processing will be more evident at the secondary elaborative phase of threat reappraisal, coping resources, and safety seeking. Given this mix of
automatic and controlled processing, one question that emerges is whether more effortful and voluntary reflection really can have a significant effect on reducing anxiety.
As previously noted, there is considerable empirical evidence from conditioning
experiments that acquired fear responses can be reduced via social transmission of
information (e.g., see discussion by Brewin, 1988). Moreover, information on the predictability and controllability of future threat, danger, or other negative events determines in large part the presence or absence of anxious apprehension (Barlow, 2002).
Furthermore, personal and clinical experience supports the assertion that conscious
controlled cognition can have a significant anxiety-reducing effect. In our everyday lives
we have all had experiences of correcting an initial feeling of anxiousness through controlled, effortful, and logical reanalysis of the perceived threat. So experimental and
anecdotal evidence is consistent with the assertion in cognitive therapy that therapeutic
interventions, like cognitive restructuring, that rely on controlled effortful thought processes can significantly contribute to anxiety reduction.
The presence of reflexive, automatic cognitive processing in anxiety does mean
that experiential or behavioral interventions, such as direct exposure to the fear stimulus, will be needed in addition to controlled cognitive interventions to reduce anxiety. Exposure-based treatment strategies are important because they enable a deeper,
more generalized and stronger activation of threat schemas and provide opportunities
to gather direct disconfirming evidence against the high threat value initially assigned
by the anxious patient (for related discussion, see Foa & Kozak, 1986). These kinds of
behavioral experiences also become powerful tools for building self- confidence in one’s
ability to deal with the anticipated threat. Chapter 6 discusses cognitive interventions at
the strategic processing level, and Chapter 7 presents various behavioral exercises used
to provide disconfirming evidence for threat.
Clinician Guideline 2.7
Strategic cognitive processing interventions and more behavioral, experiential exercises are
used to modify immediate threat appraisals and reduce heightened states of anxiety.
Self-Perpetuating Process
An anxiety episode can last from a few minutes to many hours. In fact some patients
with GAD complain that they are never really free of anxiety. So the persistence of
anxiety must be seen as a vicious cycle or a self-perpetuating process. Once the anxiety program is activated, it tends to be self-perpetuating through a number of processes. First, self-focused attention is enhanced during anxiety states so that individuals become acutely aware of their own anxiety-related thoughts and behaviors. This
The Cognitive Model of Anxiety
41
heightened attention to the symptoms of anxiety will intensify one’s subjective apprehension. Second, the presence of anxiety can impair performance in certain threatening situations, such as when the speech-anxious person goes blank or starts to perspire
profusely. Attention to these symptoms could easily interfere with the person’s ability
to deliver the speech.
In the final analysis the anxious person interprets the presence of anxiety itself as
a highly threatening development that must be reduced as quickly as possible in order
to minimize or avoid its “catastrophic effects.” In this case the person literally becomes
“anxious about being anxious.” D. M. Clark and colleagues have developed cognitive
models and interventions for panic, social phobia, and PTSD that emphasize the deleterious effects of misinterpreting the presence of anxious symptoms in a catastrophic
(or at least highly negative) manner (D. M. Clark, 1996, 2001; D. M. Clark & Ehlers,
2004). This self-perpetuating characteristic of anxiety, then, indicates that any intervention designed to interrupt the cycle must deal with any threat-related appraisals of
anxious symptoms themselves.
Clinician Guideline 2.8
Correcting misinterpretations of anxious symptoms is another important component of
cognitive therapy for anxiety disorders.
Cognitive Primacy
The cognitive model asserts that the central problem in anxiety disorders is the activation of hypervalent threat schemas that present an overly dangerous perspective on
reality and the self as weak, helpless, and vulnerable (Beck et al., 1985, 2005). From
a cognitive perspective, an initial rapid and involuntary stimulus evaluation of threat
occurs in the early phase of anxiety. It is within this framework that we view cognition
as primary in the acquisition and maintenance of fear responses. Furthermore, because
of the primacy or importance of cognition, we propose that some shift in the cognitive
conceptualization of threat is needed before any reduction in anxiety can be expected.
Without treatment, the repeated appraisal and reappraisal of threat and vulnerability
will lead to a generalization of the anxiety program so that it encompasses a broader
array of eliciting situations.
Clinician Guideline 2.9
Changing the cognitive evaluation of threat and vulnerability is necessary to reverse the generalization and persistence of anxiety.
Cognitive Vulnerability to Anxiety
There are individual differences in susceptibility or risk for anxiety disorders. Individuals are at increased risk for anxiety because of certain genetic, neurophysiological,
42
COGNITIVE THEORY AND RESEARCH ON ANXIETY
and learning histories that are causal factors in the anxiety disorders (see Chapter 1).
However, the cognitive model also asserts that particular enduring schemas involving
rules and assumptions about danger and helplessness may predispose an individual to
anxiety. See Chapter 4 for discussion of cognitive, personality, and emotional factors
that may be contributors to the etiology of anxiety.
DESCRIPTION OF THE COGNITIVE MODEL
The cognitive model depicted in Figure 2.1 is divided into an early, immediate phase of
fear response, followed by a slower, more elaborative processing phase that determines
the persistence or termination of the anxious state. Our discussion of the cognitive
model will proceed from the far left of the diagram to the end product at the far right.
Although this allows us to provide a systematic presentation of the cognitive model, in
reality all structures and processes involved in anxiety are activated almost simultaneously, and all are so interrelated that reciprocal feed forward and feedback loops are
clearly evident throughout the anxiety program.
Activating Situations, Events, and Stimuli
Environmental factors are important in the cognitive model because anxiety is a response
to an internal or external stimulus that triggers an appraisal of threat. In this sense the
model is more consistent with a diathesis– stress perspective in which particular situations or cues (the stress) activate the anxiety program in individuals with an enduring
propensity to generate primary appraisals of threat (the diathesis). Although it is possible for anxiety to occur spontaneously, as in panic attacks that occur “out of the blue,”
the more usual pattern is situation- or cue-activated anxiety.
The types of situations that can trigger anxiety are not randomly distributed. Activating situations or stimuli will differ according to the type of anxiety disorder with,
for example, social situations as relevant triggers in social phobia, stimuli that trigger
memories of a past trauma relevant to PTSD, and circumstances perceived as elevating
risk of panic attacks relevant to panic disorder. Although the situations that provoke
anxiety are personally idiosyncratic and highly diverse even within specific anxiety disorders, a stimulus will only activate the anxiety program if it is perceived as a threat to
one’s vital interests (Beck et al., 1985, 2005). This threat may be symbolic or hypothetical, as evident in GAD, or it could be perceived as real, such as when the person with
agoraphobia believes that going to a store could trigger such intense panic that a heart
attack and death could ensue.
Beck et al. (1985, 2005) conceptualized vital interests in terms of highly valued
goals or personal strivings within the social or individual domains. “Sociality” (later
termed “sociotropy”) refers to goals that involve the establishment and maintenance of
close, satisfying, and self-affirming relationships with others, whereas “individuality”
(i.e., “autonomy”) refers to goals relevant for gaining a personal sense of mastery, identity, and independence. Furthermore, these goals can be expressed in either the public
or the private sphere. From this a classification of vital interests can be constructed that
enables a better understanding of how situations might be interpreted in a threatening
manner (see Table 2.3).
The Cognitive Model of Anxiety
43
TABLE 2.3. Classification of Threats to Personal Concerns
Domain
Sociotropy
Autonomy
Public concerns
Disapproval
Disregard
Separation
Isolation
Defeat
Defection
Depreciation
Thwarting
Private concerns
Abandonment
Deprivation
Disapproval
Rejection
Disability
Malfunction
Illness
Death
Note. Based on Beck, Emery, and Greenberg (1985).
Personal strivings or goals of a social nature (sociotropy) within the public sphere
focus on our relationships within larger social settings (e.g., an audience, being in class
or at work, attending a party) that provide a sense of belonging, acceptance, approval,
and affirmation, whereas the same social strivings in the private sector refer to our more
intimate dyadic social relations (e.g., life partners, children, parents) that provide nurturance, love, empathy, and understanding. Individual personal goals within the private
sphere are concerned with achieving self-sufficiency, mastery, independence, and competence, whereas individuality (autonomy) within the public realm deals with competition and comparison where other people become instruments to achieve personal goals
and standards. Sociotropy and autonomy are understood from the perspective of the
individual, so it is the perception of acceptance, approval, independence, or competence
that is important, not some “objective” standard of whether or not a person has met his
or her goals. Also individuals will differ in the value or importance of certain strivings
for their own self-worth (for further discussion of sociotropy and autonomy, see Beck,
1983; D. A. Clark et al., 1999).
It is clear how a situation could be perceived as highly threatening if it is thought
not only to interfere or prevent the satisfaction of valued personal strivings but, even
much worse, result in a personally painful negative state of affairs (e.g., isolation, rejection, defeat, even death). For example, individuals concerned with the approval of others might feel particularly anxious if they perceive social cues of possible disapproval or
criticism in a particular social setting. On the other hand, individuals who highly value
good health and optimum functioning of their mind and body (autonomous strivings in
the private sphere) could perceive any indication of possible disease or death a serious
threat to their own survival. Any of the perceived threats common to the anxiety disorders, like loss of control or death in panic disorder and negative evaluation of others in
social phobia, can be understood in terms of threat to one’s vital interests in the public
or private spheres of sociality and autonomy.
Clinician Guideline 2.10
Determining each individual’s vital interests in the social and autonomous domains is
important for understanding development of the exaggerated personal threat evaluations
that underlie the anxiety condition.
44
COGNITIVE THEORY AND RESEARCH ON ANXIETY
Orienting Mode
Beck (1996) first proposed a cluster of schemas called the orienting mode that provides a
very rapid initial perception of a situation or stimulus. The orienting mode operates on a
matching basis such that these schemas are activated if the features of a situation match
the orienting template. The template for the orienting mode may be quite global, simply
reflecting the valence and possible personal relevance of a stimulus. That is, the orienting mode may be biased toward detection of negative and personally relevant stimuli.
We would also expect that depression and anxiety may not be differentiated at the level
of the orienting mode, with an orienting negativity bias evident in both disorders.
The orienting mode operates at the preconscious, automatic level and provides an
almost instantaneous perception of negative stimuli that could represent a possible threat
to the organism’s survival. Moreover, the orienting mode is perceptually rather than
conceptually driven. It is “an early warning detection system” that identifies stimuli and
assigns an initial processing priority. Further, attentional resources will be diverted to
situations or stimuli detected by the orienting mode. Because the function of the orienting mode is the basic survival of the organism, it is a very rapid, involuntary, and preconscious stimulus-driven registration process. At this stage stimulus detection is global
and undifferentiated, primarily identifying the valence of stimuli (negative, positive,
neutral) and its potential personal relevance. Furthermore, the orienting mode may be
biased toward detection of emotional stimuli more generally (MacLeod, 1999). Thus in
the anxiety disorders, the orienting mode is excessively tuned toward detection of negative emotional information that will subsequently be interpreted as threatening once
the primal threat mode is activated. This preconscious attentional bias means that the
anxious person has an automatic tendency to selectively attend to negative emotional
material, thus making deactivation of the anxiety program more difficult.
Primal Threat Mode Activation
The detection of possible threat-relevant negative emotional information by the orienting schemas will result in a simultaneous automatic activation of threat-related schemas
called the primal threat mode. Activation of these schemas will result in the production
of a primary threat appraisal. We use the term “primal” in this context because this
cluster of interrelated schemas is concerned with the basic evolutionary objectives of
the organism: to maximize safety and to minimize danger. For this reason the primal
threat-relevant schemas tend to be rigid, inflexible, and reflexive. They are an automatic “rapid response” system that enables the immediate detection of threat so the
organism can set about maximizing safety and minimizing danger. Once activated,
the primal threat mode tends to capture most of our attentional resources and dominates the information-processing system so that slower, more elaborative, and reflective
modes of thinking are blocked. That is, once activated, threat schemas become hypervalent and dominant, making it difficult for the anxious person to process anything but
threat. The simultaneous and immediate activation of the orienting and primal threat
schemas are evident in our previous example of the runner. Subjectively the runner
feels a sudden tension and anxiousness at hearing the dog bark. What has happened
between the dog bark and the tension is an orientation toward the sound of the dog and
the automatic primary appraisal “Could this be danger?” due to activation of primal
threat schemas.
The Cognitive Model of Anxiety
45
TABLE 2.4. Schemas of the Primal Threat Mode
Type of schema
Function
Cognitive-conceptual
Represents appraisals of threat and danger to personal
well-being, and absence or reduced likelihood of safety
Behavioral
Represents early defensive behaviors (mobilization,
immobility, escape, avoidance)
Physiological
Represents perceived autonomic arousal, physical
sensations
Motivational
Represents aims of moving away; a desire to minimize
unpredictability, lack of control, and unpleasantness
Affective
Represents subjective feelings of nervousness, agitation
The primal threat mode consists of different types of schemas all aimed at maximizing safety and minimizing danger. Table 2.4 lists the different schemas of the threat
mode and their function.
Cognitive-Conceptual Schemas
These schemas represent beliefs, rules, and assumptions that are relevant to making
inferences and interpretations of threat. Activation of the cognitive- conceptual schemas
of the primal threat mode results in the primary appraisals of threat. They enable the
selection, storage, retrieval, and interpretation of information in terms of degree of
threat to one’s vital resources. They also represent information about the self in terms
of vulnerability to threat as well as specific beliefs about the dangerousness of certain
experiences or situations in the external or internal environments.
Behavioral Schemas
Behavioral schemas consist of response disposition codes and action readiness programs
that enable a very quick and automatic early defensive response to threat. Most often this
will involve behavioral mobilization such as the fight-or-flight response regularly seen in
anxiety states. However, the behavioral schemas of the primal threat mode also enable
persons to perceive and evaluate their initial behavioral response. Behavioral responses
that are considered effective in immediately reducing threat will be reinforced and utilized on future occasions, whereas behavioral responses that do not lead to immediate
anxiety or threat reduction will tend to be discarded.
Physiological Schemas
These schemas represent information pertinent to autonomic arousal and other physical sensations. Physiological schemas are involved in the processing of proprioceptive
stimuli and allow individuals to perceive and evaluate their physiological responses (D.
A. Clark et al., 1999). Anxiety states are often associated with heightened perceptions
of physiological arousal, which can make the situation seem even more threatening. In
panic disorder the interpretation of certain physical sensations (e.g., elevated heart rate,
chest pain, breathlessness) may actually constitute the primary appraisal of threat. In
46
COGNITIVE THEORY AND RESEARCH ON ANXIETY
other anxiety disorders, like social phobia, PTSD, or OCD, perceived elevation of autonomic arousal and the physical symptoms of anxiousness can be interpreted as confirmation of threat. It is the physiological schemas of the threat mode that are responsible
for anxious persons’ threat appraisals of their heightened physical state.
Motivational Schemas
These schemas are closely related to the behavioral domain and involve representations
of our aims and intentions relevant to threat. Thus motivational schemas involve beliefs
and rules about the importance of moving away from threat or danger and of reducing
the unpredictability and aversiveness of situations. Moreover, loss of control is a state that
one is highly motivated to avoid under conditions of threat. Activation of the motivational
schemas of the primal threat mode, then, is responsible for the sense of urgency anxious
individuals feel in trying to escape or avoid a perceived threat and reduce their anxiety.
Affective Schemas
These schemas are involved in the perception of feeling states and so are integral to the
subjective experience of emotion. The affective schemas play an important functional
role in the survival of the organism by ensuring that attention is diverted to a potential threat and that some form of corrective action is taken (Beck, 1996). Activation of
threat mode affect schemas, then, produces the emotional experience individuals report
when in states of anxiety: increased nervousness, tension, agitation, feeling “on edge.”
Clinician Guideline 2.11
Utilize cognitive and behavioral interventions in cognitive therapy to reduce the accessibility and dominance of primal threat schemas, which are considered central to the experience
of anxiety.
Consequences of Threat Mode Activation
As depicted in Figure 2.1, the relatively automatic activation of the primal threat mode
sets in motion a complex psychological process that does not end simply with a primary appraisal of threat. Four additional processes can be identified that help define the
immediate fear response: increased autonomic arousal, immediate defensive and inhibitory responses, cognitive processing biases and errors, and threat-oriented automatic
thoughts and images. Each of these four processes is bidirectional with primal mode
activation responsible for their initial occurrence, but once active these processes feed
back in a manner that strengthens the primary threat appraisal.
Heightened Autonomic Arousal
Threat mode activation involves an appraisal of the heightened autonomic arousal that
characterizes anxiety states. Beck et al. (1985, 2005) stated that subjective anxiety is
The Cognitive Model of Anxiety
47
proportional to perceived estimate of danger. Thus the greater the appraised danger,
the more likely that increased autonomic arousal will be given a threatening interpretation. Highly anxious individuals often experience heightened physiological arousal as
an aversive state that confirms the initial appraisal of threat. Thus reduction of arousal
can be a prime motivation for anxious individuals. In this way a negative, threatening
interpretation of one’s increased physiological state can augment the already hypervalent threat mode.
Defensive Inhibitory Responses
Activation of the primal threat mode will lead to very rapid, reflexive self-protective
responses involving escape, avoidance (fight or flight), freeze, faint, and the like. Beck et
al. (1985, 2005) noted that these responses tend to be relatively fixed, preprogrammed,
and automatic. They are “primal” in the sense of being more innate than the complex
acquired responses associated with more elaborative processes. In the anxiety disorders,
these very immediate defensive and inhibitory responses are evident as an almost instantaneous response to a threat appraisal. For example, individuals with long-standing
OCD often report that their performance of a compulsive ritual in response to an
anxiety-provoking obsession can be so automatic that they are hardly aware of what
they are doing until they are well into the ritual. Beck et al. (1985, 2005) also recognized
that the occurrence of these protective or defensive behaviors can also reinforce primal
mode activation. They noted that these behaviors often impair performance, thus elevating the threatening nature of the situation. Thus the socially anxious individual might
automatically look away when talking to another person, which makes it more difficult
to have an engaging conversation.
Cognitive Processing Errors
Threat mode activation is “primal” in the sense that it is a relatively automatic, nonvolitional, and reflexive system for dealing with basic issues of survival. Thus one of the
cognitive by-products of this type of activation is a narrowing of attention on to the
threatening aspects of a situation. Cognitive processing, then, becomes highly selective, involving the amplification of threat and the diminished processing of safety cues.
Certain cognitive errors are evident such as minimization (underestimates the positive
aspects of personal resources), selective abstraction (primary focus on weaknesses),
magnification (views flaws as a serious shortcoming), and catastrophizing (mistakes
or threat have disastrous consequences). In anxiety these cognitive errors are manifested primarily as exaggerated estimates of the proximity, probability, and severity of
potential threat. Obviously with this type of cognitive processing dominant, the anxious
individual finds it extremely difficult to generate alternative, more constructive modes of
thinking about the situation.
Automatic Threat-Relevant Thoughts
Finally, activation of the primal threat mode will produce automatic thoughts and
images of threat and danger. These thoughts and images have an automatic quality to
48
COGNITIVE THEORY AND RESEARCH ON ANXIETY
them because they tend to be nonvolitional and intrude into the stream of consciousness. They are characterized as (1) transient or state-dependent, (2) highly specific and
discrete, (3) spontaneous and involuntary, (4) plausible, (5) consistent with one’s current emotional state, and (6) biased representation of reality (Beck, 1967, 1970, 1976).
Because automatic thoughts reflect the person’s current concerns, in anxiety disorders
they reflect themes of threat, danger, and personal vulnerability and so are hypothesized
to be content-specific to each of the anxiety disorders. In anxiety states the occurrence
of threat-relevant automatic thoughts and images will capture attention and in that way
reinforce activation of the primal threat mode.
Clinician Guideline 2.12
The adverse cognitive, behavioral, and physiological effects of threat mode activation are a
primary focus of intervention in cognitive therapy of anxiety disorders. Teach patients alternative strategies to reduce the negative impact of the threat mode.
Secondary Elaboration and Reappraisal
The quick automatic production of an immediate fear response via activation of the
primal threat mode triggers a secondary, compensatory process involving much slower,
more elaborative, and more effortful information processing. This secondary reappraisal phase always occurs with threat activation. Whether this secondary elaborative
processing leads to an increase or reduction in anxiety depends on a number of factors. The information processing that occurs at this more conscious, controlled level
will feed back into the threat mode to enhance or reduce its activation strength. In
the anxiety disorders this more constructive, reflective, and balanced thinking rarely
attains sufficient plausibility to present an alternative to primal threat mode activation. Below we discuss five cognitive phenomena associated with secondary elaborative
processing.
Evaluation of Coping Resources
A key aspect of secondary reappraisal involves the effortful evaluation of one’s ability to cope with the perceived threat. This is a strategic mode of thinking that is predominantly under voluntary and intentional control. However, in anxiety disorders the
primal threat mode activation so skews one’s elaborative thought processes that any
consideration of coping resources leads to an enhanced sense of vulnerability.
Beck et al. (1985, 2005) discussed a number of aspects of coping evaluation relevant
to anxiety. The first is a more global self-appraisal that produces self- confidence or an
increased sense of personal vulnerability. Self-confidence is “an individual’s positive
appraisal of his assets and resources in order to master problems and deal with threat”
(Beck et al., 1985, p. 68). Self- confidence will be associated with high self- efficacy and
an expectation of success (Bandura, 1977). In anxiety states, however, individuals perceive their coping resources as insufficient. A vulnerability cognitive set is reinforced,
which causes individuals to interpret incoming information in terms of their weaknesses
The Cognitive Model of Anxiety
49
rather than their strengths. A second aspect of coping evaluation concerns whether individuals believe they lack important skills to deal with the situation. The person in our
running example would experience an immediate reduction in anxiety if she recalled
previous training in dealing with dog attacks. In addition the presence of self-doubt,
uncertainty, and novel or ambiguous contexts can intensify a sense of vulnerability.
Presence of these contextual factors can mean that a cognitive set of “self- confidence”
is replaced by a “vulnerability” set (Beck et al., 1985, 2005).
One consequence of a negative evaluation of one’s coping ability is that perceived
lack of competence may cause a person to act tentatively or to withdraw from a threatening situation (Beck et al., 1985, 2005). Such tentativeness can impair one’s performance
in the situation, which only exacerbates its threatening nature (e.g., the socially anxious person trying to initiate a conversation). The anticipation of possible incompetence
and subsequent injury may inhibit approach behaviors and trigger withdrawal. This
automatic inhibition reflects a continual alteration between “confident mobility and
fearful immobility” (Beck et al., 1985, p. 73). The resulting dilemma can be described
in the following manner: “Anxiety in this instance is an unpleasant signal to stop forward progress. If the person stops or retreats, his anxiety decreases. If he advances, it
increases. If he makes a conscious decision to proceed, he may be able to override the
primal inhibitory reaction” (Beck et al., 1985, p. 72).
Clinician Guideline 2.13
Correcting maladaptive evaluations and beliefs about personal vulnerability, risk, and coping resources associated with anxious concerns is an important focus in cognitive therapy
of anxiety.
Search for Safety Cues
Beck and Clark (1997) argued that the search for safety cues is another important process that takes place at the secondary elaborative reappraisal phase. Rachman (1984a,
1984b) introduced the concept of “safety signals” to explain the discordance that can
be found between fear and avoidance (i.e., fear without avoidance and avoidance behavior in the absence of fear). Rachman proposed that in agoraphobia, for example, the
intensity of threat is primarily a function of perceived access to and speed of return to
safety. Thus the absence of reliable safety signals can leave the person in a chronic state
of anxiety, with the presence of anxiety eliciting a more vigorous search for safety cues.
The end result, however, is that the anxious person’s attempts are often ineffective,
especially in the long term. This is because safety is defined narrowly as an immediate
reduction in anxiety rather than as a long-term coping strategy. Thus the person with
panic disorder and agoraphobic avoidance might sit next to the exit in a theater, seek the
company of close friends on an outing, or carry tranquilizers as a means of procuring an
immediate sense of safety. However, all of these strategies are based on a dysfunctional
belief that “there is great danger out there and I can’t deal with it alone.” In the end
anxiety is characterized by a preoccupation with immediate safety but an unfortunate
reliance on inappropriate safety-seeking strategies.
50
COGNITIVE THEORY AND RESEARCH ON ANXIETY
Clinician Guideline 2.14
Emphasize the elimination of safety-seeking behavior in cognitive therapy of anxiety disorders.
Constructive Mode Thinking
The presence of strategic elaborative thinking provides an opportunity for more constructive, reality-based reappraisal of perceived threat. It is possible that problem-solving
strategies could be considered during secondary elaboration rather than more immediate reflexive responses aimed at self-protection or escape. Access to more realistic coping
resources is represented by schemas of the constructive mode. Constructive mode schemas are primarily acquired through life experiences and promote productive activities
aimed at increasing (not protecting) the vital resources of the individual (D. A. Clark
et al., 1999). Our ability to engage in reflective thought, to be self- conscious and evaluative of our own thoughts (i.e., metacognition), to problem-solve, and to reevaluate a
perspective based on contradictory evidence is attributable to activation of the constructive schemas.
Beck et al. (1985, 2005) proposed that anxiety is characterized by two systems, one
of which is an automatic primal inhibitory system that occurs in response to primal
threat mode activation. This system tends to be immediate and reflexive, and is aimed
at self-protection and defense. A second system, called the anxiety reduction system, is
slower, more elaborative, and processes more complete information about a situation.
The presence of anxiety can motivate a person to mobilize the more strategic processes
of anxiety reduction.
The problem in anxiety disorders, however, is that the initial automatic reflexive
(inhibitory) system activated by the primal threat mode tends to dominate information
processing and block access to more elaborative anxiety-reducing strategies represented
in the constructive schemas. Once the inhibitory system aimed at self-protection and
immediate threat reduction is activated, it is very difficult for the highly anxious person
to shift to more reflective, constructive thinking. One of the aims of cognitive therapy
is to help the anxious patient engage in more constructive mode thinking as a means of
achieving longer term reduction of anxiety.
Clinician Guideline 2.15
Encourage the development of constructive mode thinking in anxious patients to achieve
more enduring reduction in anxiety.
Initiation of Worry
Beck and Clark (1997) proposed that worry is a product of the secondary, elaborative
reappraisal process triggered by primal threat mode activation (see p. 393 for a definition of worry). In nonanxious states worry can be an adaptive process that leads to
The Cognitive Model of Anxiety
51
effective problem solving. It is anchored in constructive mode thinking in which the
individual arrives at realistic solutions based on a careful analysis of contradictory evidence. A minimal amount of anxiety may be experienced as the person considers the
possibility of negative outcomes and the consequences of ineffective coping. However,
the anxiety is not based in primal threat mode activation and so, if anything, it serves
to motivate the individual toward action.
For the highly anxious individual worry takes on pathological features that do
not lead to effective problem solving but rather to an escalation of the initial threat
appraisal. Here the worry becomes uncontrollable and almost exclusively focused on
negative, catastrophic, and threatening outcomes. Because of the domination of threat
mode thinking in the anxiety disorders, any constructive aspects of worry are blocked
and the narrow focus on negative outcomes potentiates the appraisal of threat. Thus
worry in the anxiety disorders, especially GAD, can become a self-perpetuating elaborative cycle that intensifies the anxious state and is perceived as confirmation of the
person’s initial appraisal of threat.
Clinician Guideline 2.16
Since worry is a common feature of all anxiety disorders, interventions that focus directly on
worry reduction are a major feature of cognitive therapy of anxiety.
Reappraisal of Threat
One outcome of secondary elaborative thinking is a more conscious, effortful reevaluation of the threatening situation. In nonanxious states this may result in a diminished state of anxiety as the person downgrades the probability and severity of
anticipated threat in light of contradictory evidence. Moreover, recognition of safety
features in the environment and a reappraisal of coping strategies may lead to a reduced
sense of vulnerability. In this case elaborative processing can result in a reduction in
anxiety.
In the anxiety disorders secondary elaborative thinking is dominated by the threat
mode and so is biased toward confirming the dangerousness of situations. An increased
sense of personal vulnerability is reinforced by this elaborative thinking and the realistic
safety features of the situation are overlooked. Worry and anxious rumination support
the anxious person’s initial automatic appraisal of threat. In this way, secondary elaborative cognitive processes are responsible for the persistence of anxiety, whereas primal
threat mode activation is responsible for the immediate fear response of the anxiety
program.
Clinician Guideline 2.17
Cognitive therapy seeks to help clients process disconfirming evidence that will lead to a
reevaluation of threat as less probable, severe, or imminent.
52
COGNITIVE THEORY AND RESEARCH ON ANXIETY
NORMAL AND ABNORMAL ANXIETY: A COGNITIVE PERSPECTIVE
In our description of the cognitive model, we focused primarily on pathological anxiety.
As noted earlier, fear can be adaptive and anxiety is a common experience in everyday life. So, how does the cognitive model explain the difference between normal and
abnormal anxiety? This is an important consideration for clinical practitioners as well
as researchers. After all, our goal as therapists is to normalize the experience of anxiety.
Thus what is the nature of normal cognitive processing of anxiety? Table 2.5 summarizes a few key differences at the automatic and elaborative phases of information
processing that characterize nonclinical and clinical anxiety.
Automatic Cognitive Processes in Normal Anxiety
Given the automatic and involuntary nature of the immediate fear response, it is obvious
that individuals who do not suffer an anxiety disorder have a distinct advantage over
clinical samples. In normal anxiety, the orienting mode is not as primed for the detection of negative self-referent stimuli as in the anxiety disorders. In nonclinical states,
the detection of negative stimuli will still be given attentional priority, but the range of
stimuli that would be identified as negative and potentially self-relevant would be narrower. In fact, Mogg and Bradley (1999a) reviewed evidence that less anxious individuals show attentional avoidance of low threat stimuli whereas highly anxious individuals
show enhanced attention to low, and especially moderately, threatening stimuli (see also
TABLE 2.5. Cognitive Differences between Normal and Abnormal Anxiety Predicted by the
Cognitive Model
Phase of processing
Abnormal anxiety
Normal anxiety
Orienting mode
•฀ Heightened sensitivity to negative
stimuli
•฀ More balanced sensitivity to the
detection of positive and negative
stimuli
Primal threat activation
•฀ Exaggerated primary appraisal of
threat
•฀ Negative evaluation of autonomic
arousal
•฀ Presence of threat-related
processing biases and errors
•฀ Frequent and salient automatic
thoughts and images of threat
•฀ Initiation of automatic, inhibitory
self-protective behaviors
•฀ More appropriate, reality-based
appraisal of threat
•฀ Views arousal as an uncomfortable
but not a threatening state
•฀ Attention not as narrowly focused
on threat; fewer cognitive errors
•฀ Fewer and less salient anxious
thoughts and images
•฀ Delay in inhibitory self-protective
behaviors as more elaborative
coping responses are considered
Secondary elaborative
reappraisal
•฀ Focus on weakness; low selfefficacy and negative outcome
expectancy
•฀ Poor processing of safety cues
•฀ Inaccessibility of constructive mode
thinking
•฀ Uncontrollable, threat-oriented
worry
•฀ Initial threat estimation is enhanced
•฀ Focus on strength; high self-efficacy
and positive outcome expectancy
•฀ Better processing of safety cues
•฀ Ability to access and utilize
constructive mode thinking
•฀ More controlled and reflective,
problem-oriented worry
•฀ Initial threat estimation is
diminished
The Cognitive Model of Anxiety
53
Wilson & MacLeod, 2003). Because the orienting mode in nonclinical individuals does
not show the heightened sensitivity to negative stimuli, the anxiety program is less often
activated in nonclinical than in clinical individuals.
When the anxiety program is activated in nonclinical individuals, we propose qualitative differences in primal threat mode activation compared with anxious patients. Nonclinical individuals are less likely to exhibit a preconscious attentional bias for threat,
and so their initial appraisals of threat are less exaggerated and more appropriate to the
situation at hand. In normal anxiety, threat appraisals will more accurately reflect the
consensually recognized threat value associated with internal or external situations. For
example, the panic disorder patient misinterprets chest pain as a heart attack, whereas
the nonclinical individual might interpret the chest pain as only remotely indicative of
heart disease and instead more likely due to recent strenuous physical activity.
In normal anxiety states, activation of the threat mode does not have the same negative processing effects that are evident in the anxiety disorders. For example, autonomic
arousal will be perceived as uncomfortable but not dangerous. Thus nonclinical persons
are more likely to view their aroused state as tolerable and not requiring immediate
relief. Furthermore, both automatic and more strategic attentional processes are not as
narrowly focused on threat, so nonclinically anxious individuals make fewer cognitive
errors as they process both the threatening and the nonthreatening aspects of a situation. The automatic reflexive inhibitory behaviors aimed at self-protection (fight/flight,
escape) that are so prominent in the anxiety disorders are delayed in nonclinical states.
This gives opportunity for more elaborative and strategic cognitive processes to reconsider the situation and execute a more adaptive, controlled response. The end result is
that even during times of anxiousness, nonclinical individuals will have fewer and less
salient intrusive and uncontrollable automatic thoughts and images of threat.
Secondary Elaborative Cognitive Processing in Normal Anxiety
The greatest differences between clinical and nonclinical anxiety are evident in the secondary, strategic controlled processes responsible for the persistence of anxiety. For the
clinical individual further elaboration results in a persistence and even escalation of
anxiety, whereas the same processes result in reduction and possible termination of the
anxiety program for the nonclinical person.
One of the most important differences at the elaborative phase is that nonclinical
individuals have a more balanced understanding of their personal strengths and coping
resources whereas clinical individuals tend to focus on their weaknesses and deficiencies. In nonclinical individuals this leads to high self-efficacy and expectancy of a successful or positive outcome. For individuals with anxiety disorders, negative evaluation
of their coping resources intensifies a sense of personal vulnerability and helplessness.
Second, we expect that nonclinical individuals are better able to recognize and
comprehend the safety cues in a situation compared to those with anxiety disorders.
This will allow them to arrive at a more complete understanding of their circumstances
and a more realistic assessment of its threat potential. Third, the nonclinical individual
will have greater access to constructive mode thinking so that initial threat appraisals can be reevaluated in the light of more rational, evidenced-based reasoning. In the
anxiety disorders, this type of rational, reflective thought is blocked by the hypervalent
threat schemas.
54
COGNITIVE THEORY AND RESEARCH ON ANXIETY
A fourth consideration is the quality of worry that occurs at the elaborative phase.
Normal anxiety is characterized by a more controlled, reflective, and problem-oriented
type of worry. The worry of a nonclinical person may lead to the generation of possible
solutions to a particular problem. The pathological worry in the anxiety disorders is
less controllable, more persistent, and more focused on the immediate threat of the situation. Worry in the anxiety disorders appears to intensify anxiety, whereas the worry
in nonclinical states may motivate an individual to take constructive action. The final
result is that processes at the elaborative phase may lead to diminished threat estimation
in normal anxiety, but to an intensification of the initial threat appraisal in the anxiety
disorders. In this way secondary elaborative cognitive processes are responsible for the
persistence of anxiety in abnormal states but for a controlled management and eventual
reduction of the anxiety program in normal conditions.
The cognitive perspective on normal and abnormal anxiety has direct implications
for the treatment of anxiety disorders. As cognitive therapists, our focus should be on
the elaborative strategic processes involved in secondary reappraisal. Teachman and
Woody (2004) concluded that experimental evidence supports the view that strategic
elaborative processing can override implicit or automatic cognitive processes and behavior. This is the challenge for cognitive therapists
Clinician Guideline 2.18
Shift the secondary elaborative processing and reappraisal in anxiety disorders from one of
threat enhancement to one of threat reduction, as seen in nonclinical states.
HYPOTHESES OF THE COGNITIVE MODEL
Table 2.6 presents 12 primary hypotheses derived from the cognitive model of anxiety.
Although many other hypotheses can be formulated from the cognitive perspective,
we believe these 12 hypotheses represent critical aspects of the model that provide an
empirical test of its validity. These hypotheses were derived from the central tenets of
the model (see Table 2.2) as well as the two-phase structure outlined in Figure 2.1.
Chapters 3 and 4 provide an extensive review of the empirical support for each of the
hypotheses.
SUMMARY AND CONCLUSION
It is 25 years since the cognitive model of anxiety was first introduced by Beck and colleagues (Beck et al., 1985). In this chapter we presented a reformulation of that model,
which incorporates the considerable progress made in our understanding of the cognitive contributors to the pathogenesis of anxiety. The last two decades have represented
an exceptionally productive period of cognitive- clinical research on the anxiety disorders and their treatment. In light of these developments a number of modifications,
elaborations, and clarifications were made to the cognitive model.
The Cognitive Model of Anxiety
55
TABLE 2.6. Hypotheses of the Cognitive Model of Anxiety
Hypothesis 1: Attentional threat bias
Highly anxious individuals will exhibit an automatic selective attentional bias for negative stimuli
that are relevant to threats of particular vital concerns. This automatic selective attentional threat
bias will not be present in nonanxious states.
Hypothesis 2: Diminished attentional processing of safety
Anxious individuals will exhibit an automatic attentional shift away from safety cues that are
incongruent with their dominant threat concerns, whereas nonanxious individuals will show an
automatic attentional shift toward safety cues.
Hypothesis 3: Exaggerated threat appraisals
Anxiety is characterized by an automatic evaluative process that exaggerates the threatening valence
of relevant stimuli in comparison to the actual threat valence of the stimuli. Nonanxious individuals
will automatically evaluate relevant stimuli in a less threatening manner that approximates the
actual threat value of the situation.
Hypothesis 4: Threat-biased cognitive errors
Highly anxious individuals will commit more cognitive errors while processing particular
threatening stimuli as reflected in biased estimates of the proximity, probability, and severity
of potential threat. The reverse pattern will be evident in nonanxious states where a cognitive
processing bias for nonthreat or safety cues is present.
Hypothesis 5: Negative interpretation of anxiety
Highly anxious individuals will generate more negative and threatening interpretations of their
subjective anxious feelings and symptoms than individuals experiencing low levels of anxiety.
Hypothesis 6: Elevated disorder-specific threat cognitions
Anxiety will be characterized by an elevated frequency, intensity, and duration of negative
automatic thoughts and images of selective threat and danger in comparison to nonanxious states
or other types of negative affect. Furthermore, each of the anxiety disorders is characterized by a
particular thought content relevant to its specific threat.
Hypothesis 7: Ineffective defensive strategies
Highly anxious individuals will exhibit less effective immediate defensive strategies for diminishing
anxiety and securing a sense of safety relative to individuals experiencing low levels of anxiety. In
addition highly anxious individuals will evaluate their defensive abilities in threatening situations as
less effective than nonanxious individuals.
Hypothesis 8: Facilitated threat elaboration
A selective threat bias will be evident in explicit and elaborated cognitive processes such that in
anxiety memory retrieval, outcome expectancies, and inferences to ambiguous stimuli will show a
preponderance of threat-related themes relative to nonanxious individuals.
Hypothesis 9: Inhibited safety elaboration
Explicit and controlled cognitive processes in anxiety will be characterized by an inhibitory bias of
safety information relevant to selective threats such that memory retrieval, outcome expectancies,
and judgments of ambiguous stimuli will evidence fewer themes of safety in comparison to
nonanxious individuals.
(cont.)
56
COGNITIVE THEORY AND RESEARCH ON ANXIETY
TABLE 2.6. (cont.)
Hypothesis 10: Detrimental cognitive compensatory strategies
In high anxiety worry has a greater adverse effect by enhancing threat salience, whereas worry
in low anxiety states is more likely to be associated with positive effects such as the initiation of
effective problem solving. In addition, other cognitive strategies aimed at reducing threatening
thoughts, such as thought suppression, distraction, and thought replacement, are more likely to
exhibit paradoxical effects (i.e., rebound, increased negative affect, less perceived control) in high
than in low anxious states.
Hypothesis 11: Elevated personal vulnerability
Highly anxious individuals will exhibit lower self-confidence and greater perceived helplessness in
situations relevant to their selective threats compared to nonanxious individuals.
Hypothesis 12: Enduring threat-related beliefs
Individuals vulnerable to anxiety can be distinguished from nonvulnerable persons by their
preexisting maladaptive schemas (i.e., beliefs) about particular threats or dangers and associated
personal vulnerability that remain inactive until triggered by relevant life experiences or stressors.
The present formulation places a much greater emphasis on the automatic, involuntary cognitive processes involved in the initial fear response. Although the original
cognitive model recognized that some of the mechanisms of anxiety were more innate
and automatic, the current model provides a more elaborated and fine-grained description of the automatic cognitive processes in anxiety. As the initial fear response, these
automatic processes, such as preconscious attentional threat bias, immediate threat
evaluation, and inhibitory processing of safety cues, are the catalyst for the more protracted state of anxiety that follows. Activation of threat-related schemas remains a core
feature of the cognitive model of anxiety but is now seen as responsible for maintaining an automatic threat-processing bias and its negative consequences. Thus schematic
change is still viewed as crucial to the therapeutic effectiveness of cognitive therapy for
the anxiety disorders.
Beck et al. (1985) focused much of their original discussion on the conscious, elaborative cognitive processes and structures of anxiety. The present model offers further
clarification of the role of these elaborative, strategic processes in the persistence of
anxiety. Activation of secondary, elaborative reappraisal processes, such as a conscious
evaluation of one’s coping resources, search for safety cues, attempts at more constructive or reflective thinking, and worry about and deliberate reappraisal of threat, determine the persistence of an anxious state. If a person concludes from this elaborative
processing that a significant personal threat or danger is highly probable and her ability
to establish a sense of safety through effective coping is minimal, than a state of persistent anxiety will ensue. On the other hand, anxiety will be reduced or eliminated if
the perceived probability and/or severity of threat are lowered, increased confidence in
adaptive coping is established, and a sense of personal safety is restored. Based on this
model, cognitive therapy focuses primarily on modification of these secondary, elaborative cognitive processes through specific cognitive and behavioral interventions that
shift the patient’s perspective from one of possible imminent threat to one of probable
personal safety. A change in secondary elaborative processing will reduce the propensity
The Cognitive Model of Anxiety
57
for automatic threat processing and decrease the activation threshold for threat-related
schemas.
The therapeutic strategy described in this book is theory-driven. In subsequent
chapters we discuss various cognitive restructuring and exposure-based interventions
derived from the cognitive model that can be used to modify the faulty cognitive and
behavioral processes that maintain anxiety. The basic premise is that anxiety reduction
depends on a change in the faulty cognitive processes and structures of anxiety. In the
last part of the book, a disorder-specific cognitive model and treatment protocol is proposed for each of the major anxiety disorders, which draws on the basic propositions
of the generic or “transdiagnostic” model described in this chapter. However before
considering these therapeutic applications, the next two chapters discuss the empirical
support and unresolved issues associated with our cognitive formulation for vulnerability and persistence of clinical anxiety.
Chapter 3
Empirical Status
of the Cognitive Model of Anxiety
S
ince the emergence of the cognitive model in the early 1960s (Beck, 1963, 1964,
1967), an emphasis on empirical verification has been important to its development and
elaboration. The scientific basis of the model rests on constructs and hypotheses that are
sufficiently precise to enable their support or disconfirmation in the laboratory (D. A.
Clark et al., 1999). In this chapter and the next, we present a review of the empirical status of the cognitive model of anxiety based on the 12 hypotheses presented in Table 2.6.
We begin in this chapter with the initial three hypotheses that refer to core cognitive
attributes of primal threat mode activation. The next section discusses empirical support for the cognitive, physiological, and behavioral products involved in the immediate
fear response (i.e., Hypotheses 4 to 7). The final section of this chapter reviews empirical
findings that are relevant to the persistence of anxiety (i.e., Hypotheses 8 to 10), that is,
the secondary elaboration and reappraisal phase of the model. Hypotheses 11 and 12
will be discussed in the next chapter on cognitive vulnerability to anxiety because they
deal with the etiology of anxiety.
IMMEDIATE FEAR RESPONSE: THREAT MODE ACTIVATION
Hypothesis 1. Attentional Threat Bias
Highly anxious individuals will exhibit an automatic selective attentional bias for negative
stimuli that are relevant to threats of particular vital concerns. This automatic selective
attentional threat bias will not be present in nonanxious states.
After 20 years of experimental research it is now clear that anxiety disorders are
characterized by a preconscious, automatic selective attentional bias for emotionally
threatening information (for reviews, see D. M. Clark, 1999; Macleod, 1999; Mogg &
58
Empirical Status of the Cognitive Model
59
Bradley, 1999a, 2004; Wells & Matthews, 1994; Williams et al., 1997). Because human
attentional capacity is limited, some stimuli will capture attentional resources and others will be ignored. The presence of an attentional bias for threat is expected to cause an
increased propensity to experience anxiety (McNally, 1999). Below we organized our
review of the attentional research around three types of experimentation; emotional
Stroop, dot probe detection, and stimulus identification.
Emotional Stroop
In order to experimentally investigate attentional bias in anxiety, clinical researchers
have borrowed and then modified various information-processing tasks from cognitive
experimental psychology. One of the most popular of these experimental paradigms
has been the emotional Stroop task. Based on the classic Stroop color-naming paradigm (Stroop, 1935), participants are asked to name as quickly as possible the color of
emotionally threatening (e.g., “disease,” “cancer,” “embarrassed” “disaster,” “dirty,”
“inferior”) and nonthreatening (e.g., “upward,” “network,” “leisure,” “secure”) words
printed in blue, yellow, green, or red and to disregard the meaning of the word. Typically, anxious but not nonanxious individuals take longer to name the printed color of
threat words compared with nonthreat words (e.g., Bradley, Mogg, White, & Millar,
1995; Mathews & Klug, 1993; Mathews & MacLeod, 1985; Mogg, Mathews, & Weinman, 1989; Mogg, Bradley, Williams, & Mathews, 1993). This longer color-naming
latency suggests that anxious individuals exhibit preferential allocation of attention to
the threat meaning of the word (Mogg & Bradley, 2004). Thus the extent of interference
in color-naming response by the meaning of the word is assumed to reflect attentional
bias for threat.
The emotional Stroop threat interference effect has been found in all five of the anxiety disorders discussed in this volume: panic disorder (e.g., Buckley, Blanchard, & Hickling, 2002; Lim & Kim, 2005; Lundh, Wikström, Westerlund, & Öst, 1999; McNally,
Riemann, & Kim, 1990); OCD (e.g., Kyrios & Iob, 1998; Lavy, van Oppen, & van den
Hout, 1994); social phobia (e.g., Becker, Rinck, Margraf, & Roth, 2001; Hope, Rapee,
Heimberg, & Dombeck, 1990); PTSD (e.g., J. G. Beck, Freeman, Shipherd, Hamblen,
& Lackner, 2001; Bryant & Harvey, 1995); and GAD (e.g., Bradley et al., 1995; Mogg,
Bradley, Millar, & White, 1995). Moreover, threat interference effects significantly correlate in the low to moderate range with state and symptom anxiety measures (e.g.,
MacLeod & Hagan, 1992; Mathews, Mogg, Kentish, & Eysenck, 1995; Spector, Pecknold, & Libman, 2003) and become more apparent as the threat stimulus intensity
increases from mild to severe intensity (Mogg & Bradley, 1998). In addition, the best
discrimination of attentional bias in high trait and nonclinically anxious individuals
versus low anxiety individuals might be with weak to moderately threatening cues in
which the nonanxious person would show no preferential bias for threat (Mathews &
Mackintosh, 1998).
The most consistent and robust interference effects are found with words that are
semantically related to the current emotional concerns of the anxious person (Mathews
& Klug, 1993); this content-specificity seems particularly pronounced in OCD, social
phobia, and PTSD (J. G. Beck et al., 2001; Becker et al., 2001; Buckley et al., 2002;
Foa, Ilai, McCarthy, Shoyer, & Murdock, 1993; Hope et al., 1990; Kyrios & Iob, 1998;
Lavy et al., 1994; Mattia, Heimberg, & Hope, 1993; Spector et al., 2003). However, the
60
COGNITIVE THEORY AND RESEARCH ON ANXIETY
attentional bias in GAD and, to a lesser extent, panic may be more emotionally oriented
and thus elicited by any negative emotional stimuli, and in some cases, even positive
information (e.g., Becker et al., 2001; Bradley, Mogg, White, & Millar, 1995; Buckley
et al., 2002; Lim & Kim, 2005; Lundh et al., 1999; Martin, Williams, & Clark, 1991;
McNally et al., 1994; Mogg et al., 1993; Mogg, Bradley, Millar, & White, 1995).
To investigate the automaticity of attentional threat bias, researchers modified the
emotional Stroop task to include subliminal (below conscious awareness) and supraliminal (above conscious awareness) conditions. In these studies individual threat and
nonthreat words are presented very briefly (20 milliseconds or less ) followed by a mask,
which usually involves a string of random letters presented in the same location as the
word. In some studies participants are asked to name the color of the word whereas in
other studies they are asked to name the color of the background of the word. In the
supraliminal condition the words remain unmasked on the screen until a color-naming
response is made. Figure 3.1 provides an illustration of the modified emotional Stroop
task.
In a number of studies anxious patients exhibited significantly slower color-naming
latencies to subliminal threat words, suggesting that selective attention to threat occurs
at the automatic preconscious level (e.g., Bradley et al., 1995; Kyrios & Iob, 1998; Lundh
et al., 1999; Mogg et al., 1993). Since this threat interference effect was found on both
subliminal and supraliminal trials within the same study, it suggests that attentional
bias for threat involves both automatic and elaborative cognitive processes (e.g., Bradley
et al., 1995; Lundh et al., 1999; Mogg et al., 1993).
Another important issue addressed in the emotional Stroop research is the relation
of attentional threat bias to state and trait anxiety. MacLeod and Rutherford (1992)
reported that automatic attentional threat bias is most influenced by an interaction
between state and trait anxiety. They compared nonclinical high and low trait-anxious
students on a modified emotional Stroop task and found that the high trait-anxious
students under stress (tested 1 week before exams) showed greater subliminal Stroop
interference for threat, whereas stress did not enhance threat interference for the low
trait-anxious students. In the supraliminal condition both high and low trait-anxious
students showed intentional avoidance of threat words. Other studies have also found
that increased stress and arousal are associated with greater attentional bias, especially
in high trait or fearful individuals (Chen, Lewin, & Craske, 1996; Mogg, Mathews,
Bird, & MacGregor-Morris, 1990; Richards, French, Johnson, Naparstek, & Williams,
1992; see McNally, Riemann, Louro, Lukach, & Kim, 1992, for contrary findings).
However, the effects of state and trait anxiety on attentional bias may be more complicated than first thought. High trait anxious individuals exhibit a preconscious, automatic attentional bias for threat, but unlike clinical samples, this attentional bias may be
sensitive to negative valence more generally rather than to specific threat content (e.g.,
Fox, 1993; Mogg & Marden, 1990). In addition, elevated state anxiety may lead to
greater automatic threat bias in high trait anxiety individuals (interaction effect), but at
the more elaborative, strategic level, stress may have independent effects on attentional
threat bias. MacLeod and Hagan (1992) suggested that nonclinical individuals may be
able to strategically modify their automatic threat bias, thereby eliminating any differential interference effects in the supraliminal condition. Anxious patients, on the other
hand, may fail to strategically modify their preconscious attentional threat bias so that
threat differences continue to emerge at the elaborative stage of information processing.
Empirical Status of the Cognitive Model
61
SUBLIMINAL CONDITION
Exposure time
(< 20 msec)
Masked Stimuli
Color-Naming Response
Answers
“red” for color of word
(printed in red)
(printed in red)
OR
XXXXXXXXXX
(background is red)
Answers “red” for
color of background
SUPRALIMINAL CONDITION
Exposure time
(> 500 msec)
Color-Naming Response
(printed in red)
Answers “red” for
color of word
FIGURE 3.1. Illustration of the subliminal and supraliminal conditions in a modified emotional
Stroop task.
Finally, results of an emotional Stroop experiment on PTSD led to the conclusion that
an elevation in stress or arousal might enhance automatic threat bias whereas anticipation of a more potent threat might suppress attentional bias (Constans, McCloskey,
Vasterling, Brailey, & Mathews, 2004).
There is some evidence that treatment responders do show a significant decline
in the interference effects of disorder-specific threat words whereas treatment nonresponders show no change in Stroop interference (Mathews et al., 1995; Mattia, Heimberg, & Hope, 1993; Mogg, Bradley, Millar, & White, 1995). In sum, there is consistent
evidence that preferential allocation toward threatening cues occurs at a preconscious,
automatic level of information processing in both clinically anxious and high traitanxious individuals. The emotional Stroop findings are less consistent when it comes to
demonstrating attentional biases at the slower, elaborative level of information processing.
62
COGNITIVE THEORY AND RESEARCH ON ANXIETY
Unfortunately, interpretation of the Stroop findings is hindered by limitations in its
methodology. It is possible that slower color naming could be due to diverting attention
away from threatening words rather than because of enhanced attention to the meaning
of the word (MacLeod, 1999). Also longer reaction times to threatening words could be
due to the interfering effects of an emotional reaction to the word (e.g., startle response),
or because of mental preoccupation with themes related to the word (Bögels & Mansell,
2004). Because of these potential response biases (see Mogg & Bradley, 1999a), probe
detection tests have surpassed the emotional Stroop task as the preferred experimental
paradigm for investigating attentional bias in anxiety.
Dot Probe Detection
The dot probe detection experiment is able to assess hypervigilance for threat in terms
of both facilitation and interference with dot detection without the effects of response
bias (MacLeod, Mathews, & Tata, 1986). In this task a series of word pairs is presented
so that one word is in the upper half and the other word in the lower half of a computer
screen. The trial begins with a central fixation cross presented for approximately 500
milliseconds, followed by a brief presentation (500 milliseconds) of a word pair. On
critical trials a threat and neutral word pair are presented followed by the appearance of
a dot in the location formerly occupied by one of the words. Individuals are instructed
to press a key as quickly as possible when they see the dot. Hundreds of word pair trials
are usually presented with many involving filler neutral–neutral word pairs.
A number of dot probe experiments have demonstrated an attentional threat bias
in clinically anxious patients but not in nonanxious controls. Anxious patients mainly
with a primary diagnosis of GAD exhibit significantly quicker dot probe detection after
physically and socially threatening words (MacLeod et al., 1986; Mogg, Bradley, &
Williams, 1995; Mogg, Mathews, & Eysenck, 1992). Attentional vigilance for threat
has also been found in panic disorder for detection of physically threatening words
(Mathews, Ridgeway, & Williamson, 1996), OCD for contamination words (Tata, Leibowitz, Prunty, Cameron, & Pickering, 1996), and social phobia for negative social
evaluation cues (Asmundson & Stein, 1994). Vassilopoulos (2005), however, found that
socially anxious students showed vigilance for all emotional words (positive and negative) at short exposure intervals (200 milliseconds) but avoidance of the same word stimuli at longer intervals (500 milliseconds). In addition, negative findings have also been
reported, with GAD patients failing to show attentional vigilance for threatening words
or angry faces (Gotlib, Krasnoperova, Joormann, & Yue, 2004; Mogg et al., 1991; see
also Lees, Mogg, & Bradley, 2005, for negative results with high health-anxious students).
Researchers have employed a visual dot probe task in which probe detection is
measured to pairs of pictorial stimuli involving angry versus neutral facial expressions
as a more valid representation of social evaluative threat (Mogg & Bradley, 1998).
However, visual dot probe has produced inconsistent results. While some researchers
have reported an initial selective vigilance (quicker probe detection) to angry or hostile
facial expressions at short intervals only (e.g., Mogg, Philippot, & Bradley, 2004), other
researchers failed to find vigilance for threatening or angry faces in analogue or even
clinical social anxiety groups (Gotlib, Kasch, et al., 2004; Pineles & Mineka, 2005),
and others have even reported an opposite finding, with high social anxiety character-
Empirical Status of the Cognitive Model
63
ized by a significant avoidance of emotional faces (Chen, Ehlers, Clark, & Mansell,
2002; Mansell, Clark, Ehlers, & Chen, 1999). One possibility is that social phobia
involves an initial attentional vigilance for social evaluation followed by an avoidance
of social threat stimuli once more elaborative processing occurs (Chen et al., 2002; see
findings by Mogg et al., 2004).
Dot probe experiments have been used to investigate cognitive vulnerability to
anxiety by determining if high trait anxiety is characterized by speeded detection of
threat stimuli. The most consistent finding is that high trait-anxious individuals exhibit
quicker probe detection to threatening words or faces compared to low-trait anxious
individuals, especially at shorter exposure intervals (Bradley, Mogg, Falla, & Hamilton,
1998; Mogg & Bradley, 1999b; Mogg, Bradley, Miles, & Dixon, 2004; Mogg et al.,
2000, Experiment 2). Other studies, however, have reported entirely negative findings
for trait anxiety, concluding that hypervigilance for threat was due to state anxiety (or
immediate stress) either alone or in interaction with trait anxiety (e.g., Bradley, Mogg,
& Millar, 2000; Mogg et al., 1990).
It is likely that these inconsistent findings occur because attentional bias in anxiety
involves both hypervigilance and avoidance of threat stimuli (Mathews & Mackintosh,
1998; Mogg & Bradley, 1998). Generally hypervigilance for threat has been more apparent during brief exposures when preconscious automatic processes predominate and at
higher levels of threat intensity. Avoidance of threat stimuli more likely occurs at longer
exposure intervals when more elaborative processing comes into effect and with mildly
threatening stimuli. This vigilance-avoidance pattern may be particularly evident in
specific fears, with high trait anxiety characterized by initial vigilance for threat without subsequent avoidance (Mogg et al., 2004; see Rohner, 2002, for contrary findings).
However, Rohner (2002) did not confirm this distinction between anxiety and fear.
In a study that directly examined the effects of varying levels of threat intensity,
Wilson and MacLeod (2003) compared probe detection times of high and low traitanxious students to very low, low, moderate, high, and very high anger facial expressions paired with a neutral face. All participants failed to show attentional bias to the
very low threat stimuli, attentional avoidance of mildly threatening faces, and attentional vigilance at the most intensely threatening stimuli. Interestingly, group differences in attentional deployment were only apparent with the moderately threatening
faces where only high trait-anxious group showed quicker detection of threatening than
neutral faces. Others have also found that attentional bias for threat increases with stimulus threat value (Mogg et al., 2004; Mogg et al., 2000). In a more recent study high
trait-anxious individuals showed clear evidence of facilitated attention and impaired
disengagement from high threat at 100 milliseconds but attentional avoidance at 200
or 500 milliseconds (Koster, Crombez, Verschuere, Van Damme, & Wiersema, 2006).
Finally, in an attentional training experiment by MacLeod, Rutherford, Campbell, Ebsworthy, and Holker (2002), students given training to attend away from negative words
had reduced emotional response to a stress induction compared to students trained to
attend to negative probes. This indicates that attentional bias can have a causal impact
on emotional response.
In summary both semantic (words) and visual (faces) dot probe detection research
provides the strongest experimental evidence for an automatic, preconscious hypervigilance for threat. Hypervigilance for threat is more likely when conscious elaborative
processing is restricted (shorter exposures with reduced awareness), when threat stimuli
64
COGNITIVE THEORY AND RESEARCH ON ANXIETY
match the current concerns or worries of the patient, and when threat intensity is moderate to severe. In addition facilitated attention to threat may be enhanced by an impaired
disengagement from highly threatening stimuli in anxious individuals (e.g., Koster et
al., 2006). Attentional avoidance of threat clearly plays an important role in defining
perceptual bias in anxiety but it may be less prominent in high trait anxiety (Mogg et
al., 2004). Finally, attentional bias is probably not unique to anxiety, with depression,
for example, characterized by an attentional bias for negative information (e.g., Gotlib,
Krasnoperova, et al., 2004; Mathews et al., 1996).
Stimulus Identification Tasks
Stimulus identification paradigms involve a search for threatening or nonthreatening
words within a matrix of random words or measurement of latency to identify words
presented at participants’ threshold of awareness. In a number of studies panic patients
had enhanced identification of threat stimuli (Lundh et al., 1999; Pauli et al., 1997;
see Lim & Kim, 2005, for negative findings) and social phobia individuals had facilitated identification of angry faces (Gilboa-Schechtman, Foa, & Amir, 1999). However,
studies of generalized anxiety have been more complicated, with some showing facilitated detection of threat (Mathews & MacLeod, 1986; Foa & McNally, 1986) and
others indicating that the problem might be increased distraction by threatening stimuli
(Mathews, May, Mogg, & Eysenck, 1990; Rinck, Becker, Kellerman, & Roth, 2003).
Summary
There is strong empirical support for the first hypothesis of the cognitive model. Despite
some inconsistencies across studies, there is still substantial evidence from a variety
of experimental methodologies that anxiety is characterized by a hypervigilance for
threatening stimuli and that this attentional bias is absent in low anxiety states. However, it is also clear that a number of qualifications must be added to this statement.
Attentional threat bias is more evident in the immediate or early stages of processing
when conscious awareness is reduced, when threat stimuli match the specific anxietyrelevant concerns of the individual, and when threat intensity has reached a moderate to
high level. Figure 3.2 provides a schematic illustration of how exposure duration, meaning, and threat value determine the role of selective attentional processing for threat in
anxiety (see Mogg & Bradley, 1998, 2004, for further elaboration).
Hypervigilance for threat will be absent when mildly threatening and impersonal
stimuli (e.g., general threat words) are presented at long exposure intervals. At the other
extreme, all individuals will exhibit heighten vigilance when stimuli are extremely
threatening, highly personal, and preconscious or automatic. That is, anyone will attend
to stimuli evaluated as posing a significant threat. However, it is the moderately threatening, personally specific stimuli presented at brief, preconscious exposure intervals
that will result in the exaggerated attentional threat bias that characterizes the anxiety disorders. Moderately threatening stimuli are considered threatening by vulnerable
individuals but nonthreatening to those with low anxiety (Mogg & Bradley, 1998).
However, selective attention to threat (i.e., facilitation effects) must be understood as
an interplay with avoidant (i.e., inhibitory) attentional processes, which in turn depends
Empirical Status of the Cognitive Model
Normal Attentional
Threat Bias
Moderate
High
Threat
Intensity
Exaggerated
Attentional Threat Bias
Low
Personal
Personal
Relevance
Impersonal
Delayed
Immediate
Temporality
65
Attentional Threat Bias
Absent
FIGURE 3.2. Schematic representation of threat gradient for attentional bias.
on an evaluation of stimulus threat value (Mathews & Mackintosh, 1998). An apparent
hypervigilance for threat may be due to any combination of facilitated threat detection,
impaired threat disengagement, or subsequent avoidance of threat cues with prolonged
exposure. The following clinical implication can be drawn from this research.
Clinician Guideline 3.1
Clinically anxious and vulnerable individuals automatically orient toward threat without
conscious awareness of this tendency. Some form of attentional training might help counter
this orienting bias.
Hypothesis 2. Diminished Attentional Processing of Safety
Anxious individuals will exhibit an automatic attentional shift away from safety cues that
are incongruent with their dominant threat concerns, whereas nonanxious individuals will
show an automatic attentional shift toward safety cues.
The selective attentional bias for threat reflects a narrowing of attention that accompanies emotional arousal (Barlow, 2002). “Narrowing of attention” is based on Easterbrook’s (1959) proposal that increased emotional arousal will cause a reduction in the
range of cues utilized (processed) by an organism. From an information-processing perspective, this means the higher the anxiety level, the more one’s attention will become
narrowly focused on a restricted range of mood- congruent stimuli, thereby causing a
reduction in the scope of stimulus processing (Barlow, 2002; Wells & Matthews, 1994;
see also Mathews & Mackintosh, 1998). In the present context this means that highly
anxious individuals should exhibit the greatest amount of attentional narrowing for
threat-relevant stimuli, with little attentional resources remaining to process informa-
66
COGNITIVE THEORY AND RESEARCH ON ANXIETY
tion that is mood-incongruent, such as cues of nonthreat or safety. We predict that
information signifying safety or absence of threat would be a stimulus category most
likely to be ignored in anxiety states because it is highly incongruent with this intense
focus on a narrow band of threatening information.
Two questions are relevant for this second hypothesis. First, do highly anxious individuals exhibit significantly reduced processing of relevant safety information? Second,
do nonanxious individuals show an enhanced processing bias for safety cues? Two other
issues that are related but less central to this hypothesis are whether nonanxious individuals automatically shift their attention away from threat and whether highly anxious
individuals eventually avoid threatening cues in an effort to intentionally compensate or
suppress the earlier automatic hypervigilance for threat and danger (Mathews & Mackintosh, 1998; Mogg & Bradley, 2004; Wells & Matthews, 1994).
High Anxiety: Reduced Safety Signal Processing
As noted in Chapter 2 inhibited processing of safety information is an important faulty
information-processing characteristic of anxiety. Diminished processing of safety might
be a cognitive factor that underlies the propensity of anxious individuals to engage in
safety-seeking behavior, an important factor in the persistence of anxiety (i.e., Rachman, 1984a; Salkovskis, 1996a, 1996b; Salkovskis, Clark, Hackmann, Wells, & Gelder,
1999). This is because avoidance and other safety behaviors (e.g., holding on to objects,
venturing out only when accompanied, having immediate access to medication, reassurance seeking, checking) deprive individuals of opportunities to disconfirm their catastrophic beliefs. For example, a person with panic disorder who will only go to a store
with a close family member fails to learn that she will not have a heart attack from chest
pain (i.e., the catastrophic fear belief) even though she may feel intense anxiety when
alone in the store. The catastrophic belief, then, persists despite the nonoccurrence of
heart attacks because the person engages in safety-seeking behavior (avoids stores or
takes a friend) that averts the dreaded outcome and reduces anxiety, but it also prevents
the person from learning that the belief is groundless (Salkovskis, Clark, & Gelder,
1996).
Research has shown a link between safety-seeking behavior, catastrophic beliefs, and
persistent anxiety. A questionnaire study of panic disorder (Salkovskis et al., 1996) found
evidence of the predicted associations between threat beliefs and actual safety-seeking
behavior when individuals were questioned about their responses during their most panicky or anxious episodes. In addition, brief treatment analogue studies have shown that
decreases in safety-seeking behavior lead to greater reductions in catastrophic beliefs
and anxiety (Salkovskis et al., 1999; Sloan & Tech, 2002; Wells et al., 1995). If anxious
individuals exhibit less rapid and efficient processing of safety information, this would
leave them with a narrowed intense focus on the threatening aspects of a situation. This
hypervigilance for threat combined with diminished processing of mood-incongruent
safety cues might promote more extreme and effortful attempts to reestablish a sense of
security through safety-seeking behavior (see Figure 3.3 for proposed relationships).
Only a few studies have investigated information processing of safety cues in anxiety. Mansell and D. M. Clark (1999) found that socially anxious individuals exposed
to a social-threat manipulation (give a short speech) recalled significantly fewer positive
public self-referent trait adjectives and Amir, Beard, and Prezeworski (2005) reported
Empirical Status of the Cognitive Model
HIGH ANXIETY
Early Processing
Heightened Attention to Threat
Diminished Attention to Safety
X
67
Later Processing
Delayed Threat Disengagement
Inadequate Safety Cue Processing

Increased Safety-Seeking Behavior
LOW ANXIETY
Reduced Attention to Threat
Enhanced Attention to Safety
X
Low Threat Engagement
Appropriate Safety Cue Processing

Safety-Seeking Behavior Absent
FIGURE 3.3. Proposed relation of threat and safety processing biases in high and low anxiety.
that individuals with generalized social phobia had difficulty learning nonthreat interpretations of ambiguous social information. Also, a psychophysiological study found
that combat veterans with PTSD were less expressive to emotionally positive standard
pictorial stimuli (i.e., lower zygomatic facial EMG response) after viewing a 10-minute
trauma videotape (Litz, Orsillo, Kaloupek, & Weathers, 2000; see Miller & Litz, 2004,
for failure to replicate). These findings suggest that diminished processing of nonthreat
or safety information may be evident in anxiety but this may only occur at the later
stage of strategic processing (see Derryberry & Reed, 2002). Also, the provision of
safety cues may have difficulty overriding the strong information-processing bias for
threat (i.e., Hayward, Ahmad, & Wardle, 1994) and there is even evidence that individuals with panic may show a recognition bias for “safe” facial expressions (Lundh,
Thulin, Czyzykow, & Öst, 1998).
At this point too few studies have investigated the processing of safety cues in
anxiety and so the empirical status of Hypothesis 2 cannot be determined. Clearly, studies are needed that directly compare the automatic and strategic processing of threatrelevant and safety-relevant information in clinically anxious and nonanxious controls.
In addition it would be important to establish a relationship between diminished safety
cue processing as a mediator of safety-seeking behavior.
Low Anxiety: Enhanced Safety Signal Processing
Two outcomes are possible when investigating safety signal processing in the absence of
anxiety. It is possible that attention is drawn toward positive stimuli or safety cues so
that a positivity bias is evident in nonanxious states. An alternative outcome is that no
attentional bias occurs in low anxiety so that an evenhanded processing of threatening
and safety cues prevails.
At this point we know very little about the processing of safety-relevant information in low anxiety states. In the original dot probe experiment MacLeod et al. (1986)
found that the nonanxious control group tended to shift their attention away from
threat words (see also Mogg & Bradley, 2002). However, this effect has not been replicated in most subsequent studies (e.g., Mogg, Mathews, & Eysenck, 1992; Mogg,
Bradley, et al., 2004; Mogg et al., 2000). On the other hand, MacLeod and Rutherford
(1992) found that low trait-anxious students evidenced a significant reduction in colornaming interference for threat words as their state anxiety level increased in a high stress
68
COGNITIVE THEORY AND RESEARCH ON ANXIETY
condition. Based on a color perception task, Mogg et al. (1992, Experiment 3) found
that low state anxiety individuals attended more often to manic than to neutral words.
However, in most studies the nonanxious group shows little differential result across
stimuli, suggesting an evenhanded attention to threat and nonthreat cues. Although the
key research is missing, Figure 3.3 illustrates a possible interaction between attentional
processing of threat and safety in high and low anxiety, and how these combined effects
might contribute to safety-seeking behavior in highly anxious individuals.
Threat Avoidance: An Empirical Perspective
As previously mentioned, there is emerging evidence that specific fears may be characterized by an initial vigilance for threat (at brief exposures), followed by an attentional
avoidance of threat at longer intervals, whereas high trait anxiety simply shows the
initial orientation toward threat (Amir, Foa, & Coles, 1998a; Mogg, Bradley, et al.,
2004; Vassilopoulos, 2005). However, others have reported a vigilance-avoidance pattern of attentional bias for high trait anxiety (Rohner, 2002) and increased distraction
for threat (Fox, 1994; Rinck et al., 2003). Thus questions remain about the relation
between an initial orientation to threat and subsequent disengagement followed by sustained attention away from threatening cues. It is evident that threat hypervigilance can
be countered through treatment interventions, by intentional suppression efforts, or by
creating a state of low anxiety (Mogg & Bradley, 2004). However, it is unknown how
this disengagement from threat might influence the processing of safety cues.
Summary
Empirical support for Hypothesis 2 is meager at this time because of the dearth of
relevant studies. There is some preliminary evidence that highly anxious individuals
may have diminished processing of nonthreat or safety information but this processing
bias may be evident only at the strategic and not at the automatic processing level. The
relationship between reduced safety cue processing and the occurrence of safety-seeking
behavior has not been investigated and little is known about safety signal processing in
low anxiety. Finally, mixed findings have been reported in studies on threat disengagement or avoidance, and there has been no research on its relation to safety cue processing.
Clinician Guideline 3.2
Diminished safety signal processing suggests that deliberate attentional training for safety
cues may be a useful component of anxiety treatment.
Hypothesis 3. Exaggerated Threat Appraisals
Anxiety is characterized by an automatic evaluative process that exaggerates the threatening
valence of relevant stimuli in comparison to the actual threat valence of the stimuli.
Nonanxious individuals will automatically evaluate relevant stimuli in a less threatening
manner that approximates the actual threat level of the situation.
Empirical Status of the Cognitive Model
69
There is now considerable evidence that an automatic threat appraisal process is
involved in the preattentive threat bias in anxiety. Mathews and Mackintosh (1998)
proposed that the representation of potential threat is dependent on activation of a
threat evaluation system (TES). The TES represents the threat value of a previously
encountered stimulus and is computed automatically at an early stage of information
processing. During heightened anxiety, the output from the TES increases so that a
lower threshold of stimulus intensity is required for threat valuation. Thus Mathews
and Mackintosh argue that a hypervigilant attentional threat bias occurs in response
to a prior preconscious automatic threat appraisal. Mogg and Bradley (1998, 1999a,
2004) also proposed that threat stimulus evaluation is a critical part of the automatic
information processing that occurs in anxiety (see also the self-regulatory executive
function model proposed by Wells, 2000). Recent theoretical accounts of fear and anxiety derived from a conditioning perspective propose that information is first analyzed
by feature detectors and a preconscious “significance evaluation system” that results in
a quick judgment of the fear relevance of stimuli (Öhman, 2000). Thus our contention
that automatic threat appraisal is a critical component of primal threat mode activation
is entirely consistent with other cognitive and behavioral models of fear and anxiety.
Implicit memory tasks offer an excellent experimental paradigm for investigating
the presence of automatic threat evaluation in anxiety. These tasks involve memory
retrieval in which some previously encoded information causes enhanced performance
on a subsequent task even though the individual has no awareness or recollection of the
relation between the prior experience and the task at hand (Schacter, 1990; Sternberg,
1996). In other words, previous exposure to a stimulus passively facilitates subsequent
processing of the same stimuli and this “priming effect” is thought to reflect the degree of
integrative processing that occurs during stimulus encoding (MacLeod & McLaughlin,
1995). Implicit memory more likely reflects automatic information processing, whereas
explicit memory, a deliberate and effortful retrieval of stored information, maps more
closely onto controlled, strategic processes (Williams et al., 1997).
Word Stem Completion
Implicit memory was first investigated with the word completion task. In this task individuals are presented with a list of anxiety-relevant (e.g., disease, attack, fatal) and
neutral (e.g., inflated, daily, storing) words. After a filler task, individuals are given a set
of word fragments, such as the first three letters of a word, and are asked to complete
the fragment with the first word that comes to mind. A tendency to complete the word
fragment with a less common word that was included in a previously presented word
list would be an example of implicit memory. In the following example a threat-priming
effect would be evident by completing the word fragment with a previously presented
threat word rather than with a more common neutral word.
Encoded List
Word Fragment
Possible Response
coronary
cor
coronary vs. corn
attack
att
attack vs. attend
fatal
fat
fatal vs. father
70
COGNITIVE THEORY AND RESEARCH ON ANXIETY
Studies employing word stem completion have produced mixed results that can only
be interpreted as weak evidence of implicit memory in anxiety. In some studies clinically
anxious patients or high trait-anxious individuals have generated more threat word
completions, which suggests an implicit memory for threat (e.g., Cloitre, Shear, Cancienne, & Zeitlin, 1994; Eysenck & Byrne, 1994; Mathews, Mogg, May, & Eysenck,
1989; Richards & French, 1991). However, other studies have failed to find an implicit
threat bias (e.g., Baños, Medina, & Pascual, 2001; Lundh & Öst, 1997; Rapee, McCallum, Melville, Ravenscroft, & Rodney, 1994). McNally (1995) considers word stem
completion a poor test of implicit memory in anxiety because it is strongly affected by
the physical attributes of words rather than by their meaning.
Lexical Decision Tasks
In lexical decision tasks individuals are shown a list of mixed-valence words in which
some may be anxiety-relevant, some depression-relevant, and others neutral. After a
filler task individuals are shown a second list of words that will contain some of the
“old” words, some “new” words, and also some nonword distractors (e.g., eupine,
mard, flidge). Participants are told to indicate as quickly as possible whether the stimulus is a “word” or a “nonword.” Quicker lexical decision for previously presented words
suggests an implicit memory priming effect. In anxiety we would predict quicker lexical decision for previously presented threat than nonthreat words. In this experimental
paradigm priming effects can be investigated subliminally or supraliminally depending
on whether the first exposure occurs above or below the threshold of awareness.
In two lexical decision experiments, Bradley and colleagues (Bradley, Mogg, &
Williams, 1994, 1995) failed to find evidence of an anxiety- congruent implicit memory
bias in either subliminal or supraliminal priming conditions (see also Foa, Amir, Gershuny, Molnar, & Kozak, 1997, for negative results). Amir and colleagues utilized a
more sensitive measure of automatic encoding of the meaning of information by requiring perceptual rather than word judgments to more complex stimuli. In two studies
socially anxious individuals exhibited a significant auditory or visual preferential rating
for previously presented threat stimuli that was interpreted as indicating an implicit
memory-priming effect for social threat stimuli (Amir, Bower, Briks, & Freshman,
2003; Amir, Foa, & Coles, 2000). However, Rinck and Becker (2005) failed to find an
implicit memory bias for socially threatening words in an anagram task (i.e., identify
the word from scrambled letters). Thus findings from standard lexical decision experiments or more recent perceptually oriented priming studies have not been particularly
supportive of implicit (automatic) threat evaluation in anxiety.
Primed Stimulus Identification Tasks
A number of studies have investigated implicit memory bias by determining if anxious individuals show more accurate detection of briefly presented threatening words
(stimuli) as a result of prior exposure to threat and nonthreat stimuli. MacLeod and
McLaughlin (1995) found an implicit memory bias for threat in GAD patients compared to nonanxious controls based on a tachistoscopic word identification task. The
GAD group exhibited better detection of old threat than nonthreat words, whereas
Empirical Status of the Cognitive Model
71
nonanxious controls had better identification on nonthreat than threat stimuli. However, others have failed to find speeded detection of previously presented threat versus
nonthreat words in panic disorder or PTSD (Lim & Kim, 2005; Lundh et al., 1999;
McNally & Amir, 1996). There is little evidence, then, for an implicit memory bias for
threat from stimulus identification priming studies.
Other Tests of Automatic Threat Evaluation
Amir et al. (1998a) employed a homograph paradigm to investigate activation and inhibition of threat-relevant information in individuals with generalized social phobia (GSP)
and healthy controls. Individuals read short sentences that were followed by a single
word that either did or did not fit the meaning of the sentence. Individuals had to decide
whether or not the cue word matched the meaning of the sentence. As predicted, only
the GSP group showed a slower response to cue words that followed homographs with
a possible social threat meaning. This effect was only present at short sentence priming intervals, which suggests that GSP individuals were able to suppress or inhibit an
automatic evaluation of the sentence’s threat meaning when more effortful processing
was allowed.
Employing a memory task called release of proactive interference (RPI) that taps
into the semantic organization of memory, Heinrichs and Hofmann (2004) failed to find
the predicted memory effects of socially threatening information for high socially anxious students. In fact, the opposite effect was found with the low social anxiety group
demonstrating a RPI effect for socially threatening words. In a study involving analysis
of eye movement to angry, happy, and neutral faces, Rohner (2004) was able to show
that individuals learned to avert their attention away from angry faces. In this experiment, then, anxiety was related to an implicit memory for threat avoidance.
Finally, an experimental paradigm called the Implicit Association Test (IAT) has
been used to examine automatic memory-based associations between two concepts
(Greenwald, McGhee, & Schwartz, 1998). It is considered an index of implicit attitudes because it is relatively uninfluenced by conscious controlled processes (Teachman
& Woody, 2004). In a study involving individuals highly fearful of snakes or spiders,
Teachman, Gregg, and Woody (2001) found significant differences in implicit negative
associations for snake versus spider attitudes across several semantic categories that
matched individuals’ fear concerns (Teachman & Woody, 2003; see de Jong, van den
Hout, Rietbrock, & Huijding, 2003, for negative findings of implicit associations for
spider cues in a group with high fear of spiders). Moreover, fear-related implicit associations were shown to change over the course of a three-session group exposure treatment
for phobias (Teachman & Woody, 2003).
Two studies have compared high and low socially anxious individuals on the IAT.
Tanner, Stopa, and de Houwer (2006) found that both high and low socially anxious
groups had positive implicit self-esteem as indicated by their reaction times to IAT word
classification. However, implicit self-esteem was significantly less positive in the high
social anxiety group, suggesting that a self-favoring effect was weaker in those with
high self-reported social anxiety. De Jong (2002) also concluded that high socially anxious individuals have a weaker self-favoring bias, but his results suggested this was due
to significantly higher esteem associations for others. Although only a few IAT stud-
72
COGNITIVE THEORY AND RESEARCH ON ANXIETY
ies have been published to date, they do provide some tentative support for automatic
threat association in anxiety. However, most of the studies relied on analogue samples
and so it is possible that stronger results would be found in clinical samples (Tanner et
al., 2006).
Summary
Despite consensus across various models of anxiety that some level of automatic evaluation of threat must be present in anxious states, it has been difficult to demonstrate
this effect experimentally. The few studies that are relevant to Hypothesis 3 have produced inconsistent findings. Coles and Heimberg (2002) concluded from their review
that there is modest support for implicit memory biases in all the anxiety disorders. It
may be that the results would be more supportive if priming manipulations were more
sensitive to the semantic meaning of stimuli as opposed to its perceptual properties.
It is also apparent that evidence for automatic threat bias will vary depending on the
experimental cognitive task employed. Some of the early results using the IAT suggest
that implicit associations for threat may characterize anxiety, but the results are still too
preliminary.
Clinician Guideline 3.3
The presence of automatic threat evaluation in anxiety indicates that deliberate identification, tracking and questioning of the initial threat evaluation might be helpful in diminishing the impact of automatic threat appraisals.
CONSEQUENCE OF THREAT MODE ACTIVATION
Hypothesis 4. Threat-Biased Cognitive Errors
Highly anxious individuals will commit more cognitive errors while processing particular
threatening stimuli, which will enhance the salience of threat information and diminish
the salience of incongruent safety information. The reverse pattern will be evident in
nonanxious states where a cognitive processing bias for nonthreat or safety cues is present.
Hypothesis 4 refers to the cognitive effects of fear activation that involve preconscious hypervigilance of threat, the automatic generation of threat meaning, and diminished access to safety cues. This automatic selectivity for threat will lead to further
biasing in effortful or strategic processing. We predict that threat mode activation will
lead to:
1. Overestimation of the probability, severity, and proximity of relevant threat
cues.
2. Underestimation of the presence and effectiveness of relevant safety cues.
3. The commission of cognitive processing errors such as minimization, magnification, selective abstraction, and catastrophizing.
Empirical Status of the Cognitive Model
73
Biased Threat Estimations
One of the most consistent findings in cognitive research on anxiety is that anxious individuals tend to overestimate the probability that they will encounter situations that provoke their specific anxiety state. In an early study Butler and Mathews (1983) gave clinically anxious individuals, depressed individuals, and nonclinical controls 10 ambiguous
situations. The anxious group generated significantly more threatening interpretations
and rated these negative threatening events as significantly more probable and severe
(i.e., subjective cost) than nonclinical controls but not the depressed group. This finding
was later replicated with high trait-anxious students (Butler & Mathews, 1987). Biased
estimates of threat probability have been found in subsequent research in which social
phobics overestimate the probability of experiencing negative social events (Foa, Franklin, Perry, & Herbert, 1996; Lucock & Salkovskis, 1988), claustrophobics exaggerate
the likelihood they will encounter closed spaces (Öst & Csatlos, 2000), individuals with
panic disorder interpret arousal-related scenarios and negative physical outcomes more
probable and costly (McNally & Foa, 1987; Uren, Szabó, & Lovibond, 2004), and worriers generate higher subjective probabilities for future negative events (e.g., MacLeod,
Williams, & Bekerian, 1991). In this latter study increased access to reasons why the
negative event would happen and reduced access to why it would not happen (i.e., safety
features) predicted probability judgments.
Cognitive bias should be most evident during fear activation. The positive correlation between heightened probability or severity (i.e., costly) estimates of threat and
intensity of anxious symptoms is consistent with this prediction (e.g., Foa et al., 1996;
Lucock & Salkovskis, 1988; Muris & van der Heiden, 2006; Öst & Csatlos, 2000;
Woods, Frost, & Steketee, 2002). Moreover, causal relations between anxiety and threat
perception have been found in fear provocation experiments. In various studies anxious
and phobic individuals predict that they will experience more fear and panic attacks
than actually happens when exposed to the fear situation (e.g., Rachman, Levitt, &
Lopatka, 1988b; Rachman & Lopatka, 1986; Rachman, Lopatka, & Levitt, 1988).
This tendency to overestimate the likelihood of threat has also been found in the worry
concerns of chronic worriers (Vasey & Borkovec, 1992) and in the exaggerated negative
appraisals of social performance generated by socially anxious individuals (Mellings
& Alden, 2000; Stopa & Clark, 1993). However, with repeated experience, individuals show a decrease in their overpredictions of fear so that their estimates more closely
match their actual fear level.
Maladaptive Looming Effect
Along with exaggerated estimates of threat probability and severity, inaccurate appraisals of the proximity of danger are also an aspect of biased cognitive processing in anxiety. Riskind and Williams (2006) emphasize that “mental representations of dynamically
intensifying danger and rapidly rising risk” (pp. 178–179), termed looming maladaptive
style, are a key component of threat appraisal in anxiety. According to Riskind and colleagues, a critical feature of any threatening stimulus is the perception of the threat as
moving and intensifying in relation to the self in terms of physical or temporal proximity of real events but also in terms of the mental rehearsal of the potential time course
of future events (Riskind, 1997; Riskind, Williams, Gessner, Chrosniak, & Cortina,
74
COGNITIVE THEORY AND RESEARCH ON ANXIETY
2000). Exaggerated threat in anxiety must be understood in terms of this dynamic
danger content involving qualities such as the velocity (directional speed), accelerating
momentum (rate of increase), and direction (coming toward the person) of threat (Riskind, 1997; Riskind & Williams, 1999, 2005, 2006). The looming vulnerability model,
then, maintains that anxiety occurs when threat is appraised as rapidly approaching
or developing such as an approaching snake, a deadline, an illness, or a social failure
(Riskind, 1997). It is considered a key feature of the danger schema activated in anxiety
and so is a specific construct that is applicable to all anxiety states from simple phobias
to more abstract phenomena like worry and GAD (Riskind & Williams, 1999).
Riskind and Williams (2006) review an emerging research literature that supports
the role of perceived intensifying danger and rapidly rising risk (i.e., looming) in predicting other features of anxious phenomenology. Experimental studies indicate that
moving fear stimuli (e.g., videotape of tarantulas) elicit more fear and threat-related
cognitions than stationary fear or neutral stimuli (Dorfan & Woody, 2006; Riskind,
Kelly, Harman, Moore, & Gaines, 1992) and phobic anxiety is associated with a greater
tendency to perceive a fear stimulus (e.g., spider) as changing or moving rapidly toward
one’s self (e.g., Riskind et al., 1992; Riskind, Moore, & Bowley, 1995; Riskind & Maddux, 1993). In addition, the Looming Maladaptive Style Questionnaire (LMSQ), which
assesses a tendency to generate mental scenarios that involve movement toward some
dreaded outcome, is uniquely associated with several features of anxious phenomenology (Riskind et al., 2000) and may be a latent common factor that underlies OCD,
PTSD, GAD, social phobia, and specific phobias (Williams, Shahar, Riskind, & Joiner,
2005). Overall these findings are consistent with the observation that anxious individuals misjudge the impending nature of threatening stimuli, leading them to the erroneous conclusion that danger is closer or more immediate than is actually true. Riskind’s
research indicates this heightened sensitivity to the kinetic qualities of danger is an
important aspect of biased threat appraisals in anxiety.
Cognitive Errors
Surprisingly little research has investigated the relevance of depressive cognitive errors
(e.g., dichotomous thinking, overgeneralization, selective abstraction) for anxiety. In a
study of thought content individuals with GAD generated more imperatives (“have to/
should”) and catastrophizing words than dysphoric and nonanxious students, and all
participants generated more cognitive errors during worry than during a neutral condition (Molina, Borkovec, Peasley, & Person, 1998). Despite a paucity of research, it is
likely that anxious individuals do exhibit many of the same cognitive errors found in
depression, especially when dealing with information relevant to their fear concerns.
However, research is needed to determine the role of inferential cognitive errors in the
anxiety disorders.
Summary
We began our review of Hypothesis 4 with three predictions concerning the role of
cognitive errors in fear activation. Unfortunately, only one of these predictions has been
tested empirically. The empirical evidence is consistent in showing that anxious individuals exaggerate the likelihood and probably the severity of negative situations related
Empirical Status of the Cognitive Model
75
to their anxious concerns. This cognitive bias for threat estimation appears relevant to
most of the anxiety disorders, although it is still debatable whether it is specific only to
anxiety. The research on looming cognitive style clearly indicates that overestimating
the proximity or impending nature of danger is a critical aspect of biased threat evaluation that potentiates the anxious state.
It is likely that highly anxious individuals generate the same types of cognitive
errors that we see in depression. Catastrophizing is well known in panic disorder, but it
is likely that dichotomous thinking, selective abstraction, magnification/minimization,
overgeneralization, and other forms of rigid and absolutistic thinking are prominent
in all the anxiety disorders. Research is needed to determine whether some of these
cognitive errors are specific to anxiety-relevant concerns and what role they play in the
persistence of fear activation. It would also be helpful to move beyond static paper-andpencil measures of cognitive errors to “online assessment” of thought content during
fear provocation.
At this time we have no information on the role of cognitive errors in the diminished processing of safety cues that is considered an important feature of fear activation.
We assume that if cognitive processing errors can lead to an overestimation of threat,
then this same cognitive processing style might lead to an underestimation of safety.
This latter proposition, however, must await empirical investigation.
Clinician Guideline 3.4
Repeated experiences with situations involving varying levels of impending threat that disconfirm anxious individuals’ exaggerated threat expectancies are critical in modifying the
erroneous thinking style that contributes to the persistence of the anxious state.
Hypothesis 5. Negative Interpretation of Anxiety
Highly anxious individuals will generate more negative and threatening interpretations of
their subjective anxious feelings and symptoms than individuals experiencing low levels of
anxiety.
In the cognitive model (see Figure 2.1) increased autonomic or physiological arousal
is another prominent feature of threat mode activation. Hypothesis 5, however, refers to
the cognitive processes associated with physiological arousal. It is proposed that highly
anxious individuals will perceive their heightened arousal, anxious feelings, and other
somatic symptoms of anxiety as more threatening and unacceptable than low anxious
individuals. It is also expected that this “fear of fear” (Chambless & Gracely, 1989) will
be more evident during highly anxious states and will motivate individuals to terminate
the fear program.
Beck et al. (1985, 2005) identified another aspect of this negative interpretation of
anxiety, “emotional reasoning,” in which the state of feeling anxious is itself interpreted
as evidence that danger must be present. Later Arntz, Rauer, and van den Hout (1995)
referred to this as “ex-consequentia reasoning” which involves the fallacious proposition “If I feel anxious, there must be danger” (p. 917). They found that spider phobic,
panic, social phobic, and other anxiety disorder patients but not nonclinical controls
76
COGNITIVE THEORY AND RESEARCH ON ANXIETY
were significantly influenced in their danger ratings of hypothetical anxiety scripts by
the presence of anxiety response information.
It is proposed that different aspects of the subjective experience of anxiety will be
perceived as threatening depending on the nature of the anxiety disorder. In some cases
it will be the physiological symptoms that are considered most unacceptable, whereas in
other disorders it is cognitive phenomena (i.e., worry or unwanted intrusive thoughts)
or even the heightened sense of general anxiousness that is perceived as most disturbing.
Whatever the actual focus, it is the state of being anxious that is considered threatening
and intolerable to the person. Table 3.1 presents the specific negative interpretations of
anxiety associated with each of the anxiety disorders discussed in this volume.
Empirical Evidence
Negative interpretation of physiological arousal is a central process in the cognitive model
of panic disorder (see Chapter 8 for further discussion). Questionnaire studies indicate
that individuals with panic disorder are more likely to negatively (even catastrophically)
misinterpret bodily sensations associated with anxiety and to report more distress when
experiencing these symptoms than nonclinical individuals or those with other types
of anxiety disorders (e.g., D. M. Clark et al., 1997; Harvey, Richards, Dziadosz, &
Swindell, 1993; Hochn-Saric, McLeod, Funderburk, & Kowalski, 2004; Kamieniecki,
Wade, & Tsourtos, 1997; McNally & Foa, 1987; Rapee, Ancis, & Barlow, 1988). Also,
experimental research indicates that panic patients are more likely to feel anxious or
even to panic when they focus on induced or naturally occurring bodily sensations
(Antony, Ledley, Liss, & Swinson, 2006; Pauli, Marquardt, Hartl, Nutzinger, Hölzl, &
Strain, 1991; Rachman, Lopatka, & Levitt, 1988; Rachman, Levitt, & Lopatka, 1988;
Hochn-Saric et al., 2004). Together these studies provide a strong empirical basis that a
heightened misinterpretation of physiological arousal is a key process in panic.
For individuals with GAD a focus on the more cognitive symptoms of anxiety will
characterize their negative interpretation of anxiousness. Adrian Wells first noted that
“worry about worry” (i.e., metaworry) is a prominent feature of GAD that distinguishes
high worriers from those who are nonworriers (Wells, 1997; Wells & Butler, 1997;
TABLE 3.1. Specific Negative Interpretations of Anxiety Associated with Each
of the Anxiety Disorders
Anxiety disorder
Focus of negative interpretation of anxiety
Panic disorder
Physiological arousal, specific bodily sensations
Generalized anxiety disorder
Subjective experience of worry (“worry about worry”)
Social phobia
Somatic and behavioral indicators of being anxious in social settings
Obsessive–compulsive disorder
Anxious feeling associated with certain unwanted intrusive thoughts,
images, or impulses
Posttraumatic stress disorder
Specific physiological and emotional arousal symptoms associated
with trauma-related mental intrusions
Empirical Status of the Cognitive Model
77
Wells & Mathews, 1994). Metaworry involves a subjective negative appraisal of the
significance, increased incidence, and perceived difficulties associated with the uncontrollability of worry (Wells & Mathews, 1994). Evidence that GAD is associated with
heightened metaworry would support Hypothesis 5 and indicate that in generalized
anxiety a negative interpretation of the act of worrying (e.g., “If I don’t stop worrying,
I’ll end up an emotional wreck”) contributes to an intensification and persistence of the
anxious state. In fact, various studies have shown that GAD patients were distinguished
from patients with other anxiety disorders (especially social phobia) by heightened
scores on metaworry (Wells & Carter, 2001) and there is a strong relationship between
metaworry and increased tendency to experience pathological worry (Wells & Carter,
1999; Wells & Papageorgiou, 1998a; see also Rassin, Merchelback, Muris, & Spaan,
1999). An early study by Ingram (1990) found that generalized anxiety and depression
were characterized by a heightened focus on one’s thoughts, sensations, and feelings
as indicated by Fenigstein, Scheier, and Buss’s (1975) Self- Consciousness Scale (SCS).
These studies are consistent with Hypothesis 5, indicating that an increased focus on the
negative characteristics of worry will exacerbate the general anxiety state.
In social phobia negative interpretation of anxious symptoms in social situations
because of a concern that anxiety will be perceived negatively by others is a central
feature of the disorder (see D. M. Clark & Wells, 1995; Wells & Clark, 1997). Various
studies have found that social phobia is characterized by negative appraisal of anxietyrelated interoceptive cues that leads to erroneous inferences about how one appears to
others and subsequently to heightened subjective anxiety (for review, see D. M. Clark,
1999; Bögels & Mansell, 2004). Elevated self-focused attention has been found in social
anxiety (e.g., Daly, Vangelisti, & Lawrence, 1989; Hackman, Surawy, & Clark, 1998;
Mellings & Alden, 2000). Moreover, a specific focus on anxious symptoms (e.g., blushing) intensifies anxiety in high social anxiety but not in low social anxiety (Bögels &
Lamers, 2002; see Bögels, Rijsemus, & De Jong, 2002, for contrary findings).
Experimental research has also supported the cognitive model. Mansell and D. M.
Clark (1999) found a significant association in high but not low social anxiety between
perception of bodily sensations and ratings of how anxious individuals thought they
appeared to others. Mauss, Wilhelm, and Gross (2004) compared high and low socially
anxious students before, during, and after a 3-minute impromptu speech and found that
the high social anxiety group perceived a greater level of physiological arousal, felt more
anxious, and exhibited more anxious behavior than the low anxious group even though
there were no significant group differences in actual physiological activation. Moreover,
self-reported anxiety correlated with perceived but not actual physiological activation
for the total sample. These findings are consistent with Hypothesis 5. Social phobia is
characterized by a heightened focus on anxious symptoms that clearly intensifies the
anxious state.
In cognitive accounts of OCD the central problem is the faulty appraisal of
unwanted intrusive thoughts, images, or impulses of dirt, contamination, doubt, sex,
causing injury to others, and the like (D. A. Clark, 2004; Salkovskis, 1989, 1999; Rachman, 1997, 1998, 2003). Thus obsessional thinking develops when an unwanted intrusive thought, image, or impulse is misinterpreted as representing a significant potential
threat to one’s self or others and the person perceives a heightened sense of personal
responsibility to prevent this anticipated threat. Rachman (1998) suggested that “emo-
78
COGNITIVE THEORY AND RESEARCH ON ANXIETY
tional reasoning” could play an important role in the faulty appraisal of obsessional
intrusions. Any anxiety associated with an intrusion could be misinterpreted as confirming the significance and potential dangerousness of the thought. This would be an
example of “ex-consequentia reasoning” (Arntz et al., 1995) contributing to the faulty
appraisal and escalation of the intrusion (e.g., “If I feel anxious by the thought of being
dirty and potentially contaminating others, then I must be in danger of infecting others.”).
There is a strong association between the subjective anxiousness or emotional distress of an intrusive thought, and its frequency, uncontrollability, and obsessionality (e.g.,
Freeston, Ladouceur, Thibodeau, & Gagnon, 1992; Parkinson & Rachman, 1981a; Purdon & Clark, 1993, 1994b; Salkovskis & Harrison, 1984). Moreover, individuals with
OCD rate their obsessions and other unwanted intrusions as more anxiety-provoking
than do nonobsessional controls (Calamari & Janeck, 1997; Janeck & Calamari, 1999;
Rachman & de Silva, 1978). In a diary study involving 28 patients with OCD the individual’s most upsetting obsession was rated as more frequent and more meaningful in
terms of importance and control of thought than the least upsetting obsessions (Rowa,
Purdon, Summerfeldt, & Antony, 2005). These findings are consistent with the view
that OCD is characterized by a heightened sensitivity to certain OCD-related mental
intrusions that may in part be due to the anxiety- eliciting properties of the obsession.
However, research is needed that specifically investigates whether OCD is characterized
by a misinterpretation of anxious feelings associated with obsessional intrusions and
that this, in turn, contributes to a heightened state of general anxiousness.
Negative interpretation of anxious symptoms associated with trauma-related intrusions is a key process emphasized in cognitive theories of PTSD (Brewin & Holmes,
2003; Ehlers & Clark, 2000; Wells, 2000). Many studies have now shown that negative
interpretation of initial PTSD symptoms plays a causal role in the persistence of PTSD
(see review by Brewin & Holmes, 2003). In addition, negative appraisal of unwanted
trauma-relevant intrusive thoughts or images is predictive of the severity and persistence
of PTSD (Halligan, Michael, Clark, & Ehlers, 2003; Steil & Ehlers, 2000; Mayou,
Bryant, & Ehlers, 2001). These findings, then, are entirely consistent with Hypothesis
5, indicating that negative and threatening interpretations of trauma-related anxious
symptoms contribute significantly to the persistence of PTSD.
Summary
This brief review of the empirical research on enhanced negativity bias in the interpretation of anxious symptoms indicates strong empirical support for Hypothesis 5. Research
spanning all five anxiety disorders found evidence that enhanced negative interpretation
of anxiety or “fear of fear” was a contributor to the persistence of anxiety (see also
chapter 4 on the related concept of anxiety sensitivity). Panic disorder is characterized
by threat misinterpretations of the physical symptoms of anxiety, GAD by metaworry,
social phobia by heightened self-focused attention on internal states of anxiousness,
OCD by the anxiety-arousing properties of mental intrusions, and PTSD by physiological arousal elicited by trauma-related internal and external triggers. In each case a tendency to perceive anxiety itself in a threatening manner contributed to the persistence
of the unwanted emotional state.
Empirical Status of the Cognitive Model
79
Clinician Guideline 3.5
The idiosyncratic meaning of anxious symptoms (i.e., the significance of heightened anxiousness) must be assessed and treated with cognitive restructuring as part of the intervention for reducing primal threat mode activation.
Hypothesis 6. Elevated Disorder-Specific Threat Cognitions
Anxiety will be characterized by an elevated frequency, intensity, and duration of negative
automatic thoughts and images of selective threat and danger in comparison to nonanxious
states or other types of negative affect. Furthermore, each of the anxiety disorders is
characterized by a particular thought content relevant to its specific threat concerns.
One of the conscious phenomenal manifestations of primal threat mode activation
is the frequent and repeated intrusion into conscious awareness of automatic thoughts
and images related to the specific fear concerns of the individual. There is, in fact, a
very large empirical literature that has demonstrated a preponderance of harm, threat,
and danger cognitions and images in panic disorder (Argyle, 1988; McNally, Hornig,
& Donnell, 1995; Ottaviani & Beck, 1987); GAD (Beck, Laude, & Bohnert, 1974;
Hibbert, 1984); social phobia (Beidel, Turner, & Dancu, 1985; Hackmann et al., 1998;
Turner, Beidel, & Larkin, 1986); and OCD (Calamari & Janeck, 1997; Janeck & Calamari, 1999; Rachman & de Silva, 1978; Rowa et al., 2005); as well as posttraumarelevant threatening intrusions in PTSD (Dunmore, Clark, & Ehlers, 1999; Mayou et
al., 2001; Qin et al., 2003; Steil & Ehlers, 2000). This “softer version” of Hypothesis 6,
then, has been well documented in the empirical literature.
The more controversial aspect of Hypothesis 6 is the “strong version” predicting
that each of the anxiety disorders will show a specific cognitive profile, and that this
profile will distinguish anxiety from other negative emotional states. Table 3.2 presents
the automatic thought content that characterizes each of the anxiety disorders.
There are two aspects to the “specificity” question in this hypothesis. First, to what
extent is anxiety distinguishable from depression, with the former characterized by
thoughts of harm and danger whereas the latter is distinguished by thoughts of loss and
failure? And second, is there a specific cognitive profile that characterizes each of the
anxiety disorder subtypes?
Cognitive Specificity: Distinguishing Anxiety from Depression
The content-specificity hypothesis states that “each psychological disorder has a distinct
cognitive profile that is evident in the content and orientation of the negative cognitions and processing bias associated with the disorder” (Clark et al., 1999, p. 115). The
content or orientation of the automatic thoughts and processing bias that characterizes
anxiety states focuses on the possibility of future physical or psychological threat/danger and the sense of increased personal vulnerability or lack of safety. In depression the
predominant cognitive theme concerns past personal loss or deprivation. In fact, global
hopelessness as well as hopelessness about specific life problems is significantly greater
in major depression than in GAD (Beck, Wenzel, Riskind, Brown, & Steer, 2006). The
80
COGNITIVE THEORY AND RESEARCH ON ANXIETY
TABLE 3.2. Types of Automatic Thoughts and Images That Characterize Specific
Anxiety Disorders
Anxiety disorder
Thematic content of automatic thought/image
Panic with/without agoraphobic
avoidance
. . . of physical catastrophe (e.g., fainting, heart attack,
dying, going crazy)
Generalized anxiety disorder
. . . of possible future loss and failure in valued life domains
as well as fear of losing control or inability to cope
Social phobia
. . . of negative evaluation by others, humiliation, poor
social performance
Obsessive–compulsive disorder
. . . of losing mental or behavioral control that results in
serious harm to self or others.
Posttraumatic stress disorder
. . . of past trauma and its sequelae
cognitive model, then, asserts that anxiety and depression can be distinguished by the
content (and temporal orientation) of the negative automatic thoughts and interpretations generated by the individual.
In our own studies future-oriented threat-related cognitions distinguished panic
and GAD from major depression/dysthymia (Clark, Beck, & Beck, 1994) and threatrelated cognitions showed a closer, more specific relation with an anxiety than a depression symptom dimension (Clark, Beck, & Stewart, 1990; Clark, Steer, Beck, & Snow,
1996). These findings have been supported in other studies, although anxious cognitions
appear to have a greater degree of nonspecificity than depressive cognitions (e.g., Beck,
Brown, Steer, Eidelson, & Riskind, 1987; Ingram, Kendall, Smith, Donnell, & Ronan,
1987; Jolly & Dykman, 1994; Jolly & Kramer, 1994; Jolly, Dyck, Kramer, & Wherry,
1994; Schniering & Rapee, 2004). In a meta-analysis of 13 studies, R. Beck and Perkins (2001) found only partial support for the content-specificity hypothesis. Anxious
and depressive cognition measures were significantly correlated with both their corresponding and noncorresponding mood/symptom measures and the cognition measures
showed an average correlation of .66 with each other. Yet, quantitative comparisons did
reveal that the depressive cognition measures had significantly higher correlations with
depression than with anxious symptoms, but the anxious cognitions were equally correlated with depression and anxiety. The authors concluded that threat-related cognitions
may not have the same degree of specificity as depressive cognitions (R. Beck & Perkins,
2001; see similar conclusion reached in review by Clark et al., 1999), although certain
clinical populations or levels of symptom severity may show more or less specificity
(Clark et al., 1996; Ambrose & Rholes, 1993).
The apparent lack of specificity for anxious cognitions may reflect a greater degree
of heterogeneity for anxious than depressive cognitions. R. Beck and Perkins (2001)
suggest two possibilities for the lack of specificity with anxious cognitions. Is it possible
that a subset of anxious thought can be identified that is specific to particular anxiety
disorders, whereas other types of anxious thinking may be more generally related to
anxiety and depression? Or depressive cognitions may show greater specificity because
they are related to low positive affect, which is a specific mood–personality construct
of depression, and anxious cognition is less specific because it is the cognitive face of
Empirical Status of the Cognitive Model
81
high negative affect, which is a mood– personality dimension common to all emotional
disorders.
There is evidence that specificity may only apply to a subset of anxious cognitions.
Jolly and Dykman (1994) reported that some threat cognitions were more related to a
general negativity factor, whereas other cognitions related to physical or health-related
threat were more specific to anxiety. In other research anxious overconcern (i.e., worry)
emerged as a common feature of all anxiety disorders, whereas negative evaluation of
others or social threat may evidence more subtype-specificity (Becker, Namour, Zayfert,
& Hegel, 2001; Mizes, Landolf-Fritsche, & Grossman-McKee, 1987). Finally, Riskind
(1997) has argued that looming vulnerability, the perception of threat movement, may
offer better precision in distinguishing anxiety from depression because it incorporates
time and rate of change in its conceptualization of threat appraisal. Although still tentative, it appears that only certain types of threat-related cognitions such as concerns
about physical symptoms, health, social evaluation, and impending danger are specific
to anxiety, whereas anxious apprehension or worry may be more evident in both anxiety and depression.
Cognitive Specificity in Anxiety Disorder Subtypes
Less research has investigated whether a specific cognitive content is associated with the
anxiety disorder subtypes. In two studies R. Beck and colleagues found that worry was
common to anxiety and depression and a strong predictor of negative affect, whereas
hopelessness was predictive of low positive affect and panic-related cognitions were
clearly specific to anxiety states (R. Beck, Benedict, & Winkler, 2003; R. Beck et al.,
2001). In a confirmatory factor analysis of self-reported anxious and depressive selfstatements, self-statements reflecting depression/hopelessness and self-statements reflecting anxiety/uncertainty about the future had large and significant loadings on a general
negativity factor (Safren et al., 2000).
One of the most direct tests of cognitive content-specificity among anxiety disorder
subtypes was reported by Woody, Taylor, McLean, and Koch (1998). They found that
patients with panic disorder scored significantly higher on a measure of threat-related
cognitions that were unique to panic (i.e., the UBC Cognitions Inventory—Panic subscale) compared to patients with major depression. However, the two groups did not
differ on the Cognitions Checklist—Anxiety subscale, which the authors claim assesses
more general conceptions of anxious cognitions.
Summary
Over the years numerous studies have shown that automatic thoughts and images of
threat, danger, and harm occur with greater frequency and intensity in the anxiety disorders when fear is activated. Consequently there is ample evidence supporting the basic
assertion of Hypothesis 6. Whether thoughts of threat and danger are a specific marker
of anxiety has been more equivocal, and whether each anxiety disorder has its own
unique cognitive content that distinguishes it from other emotional states has not been
subjected to adequate empirical investigation. However, a number of tentative conclusions can be drawn about cognitive content-specificity in anxiety. It is likely that only
some forms of anxious thought will show the level of specificity predicted by Hypothesis
82
COGNITIVE THEORY AND RESEARCH ON ANXIETY
6. Specificity is more likely when researchers focus on thought content that characterizes each of the disorder subtypes (see Table 3.2) rather than more general forms of
apprehensive thought. Moreover, cognitive content-specificity may be more apparent at
higher levels of symptom severity or in clinical groups that present with greater diagnostic homogeneity (e.g., pure anxiety disorder groups). Failure to find specificity in
the anxiety disorders could reflect the inadequacies of the measures employed, especially if self-report questionnaires are used that underrepresent the more specific forms
of cognition associated with the anxiety subtypes. Also, the high rate of comorbidity
between anxiety and depression has complicated efforts to investigate level of specificity
in pathognomonic processes. Cognitive specificity research would be advanced if investigators compared “pure” (single-diagnosis) anxiety and depression groups using specialized instruments of negative thought content. Until then, much remains unknown
about the parameters of cognitive content-specificity in anxiety.
Clinician Guideline 3.6
Clinicians should use thought records, diaries, and other self-monitoring forms to obtain a
“real-time” assessment of the automatic thought and image content that intrudes into conscious awareness during fear activation. Specific themes of threat and danger will provide
valuable diagnostic and assessment information for constructing a case formulation of the
anxiety disorder.
Hypothesis 7. Ineffective Defensive Strategies
Highly anxious individuals will exhibit less effective immediate defensive strategies for
diminishing anxiety and securing a sense of safety relative to individuals experiencing
low levels of anxiety. In addition, highly anxious individuals will evaluate their defensive
abilities in threatening situations as less effective than nonanxious individuals.
Hypothesis 7 focuses on the final consequence of threat mode activation (see Figure
2.1). It is proposed that fear activation involves an automatic defensive response that
is aimed at immediate reduction or avoidance of fear and the reinstatement of safety.
This rapid response system is not an effortful intentional coping response but instead a
fundamental biologically based adaptational system that is triggered when the organism encounters a potentially life-threatening situation (Öhman & Mineka, 2001). The
adaptational value of fear is partly due to its ability to trigger an immediate defensive
response.
Fear has evolved to deal with situations involving physical danger that are potentially life-threatening and so primitive alarm reactions may be effective for external
dangers. However, they are less useful, even counterproductive, for the more abstract,
protracted, and internally oriented threats that characterize the anxiety disorders. Beck
et al. (1985, 2005) proposed that two automatic behavioral defensive systems can be triggered by threat. The first is an active, energic system involving mobilization (e.g., fight,
flight) in response to danger. The second is a more passive, anergic system that involves a
stereotypic immobility response (e.g., fainting). Craske (2003) presented a threat immi-
Empirical Status of the Cognitive Model
83
Automatic Behavioral Response
Avoidance
Escape
Immobility
Threat Mode
Activation
Automatic Cognitive Response
Attentional avoidance
Distraction
Thought Suppression
Automatic Safety Seeking
Active responses to restore
personal safety
FIGURE 3.4. The automatic defensive response system associated with threat mode activation.
nence model in which increased proximity and detection of a threat is associated with a
corresponding state of autonomic arousal in preparation for fight or flight.
Figure 3.4 summarizes the behavioral, cognitive, and safety-seeking processes
involved in the automatic defensive reaction elicited by threat mode activation.
Behavioral Escape and Avoidance
Escape and avoidance behavior is so prominent in anxiety states that it is included as
one of the cardinal DSM-IV diagnostic features of social phobia, PTSD, specific phobia,
and panic disorder (APA, 2000). Furthermore, attempts to ignore, suppress, or neutralize obsessions in OCD and the ineffective control of worry in GAD can be considered
examples of escape responses in these disorders. Escape and avoidance responses are
so closely associated with subjective fear that their occurrence is taken as an important
marker of fear expression (Barlow, 2002).
Behavioral, biological, and emotion theories of fear are almost universal in their
agreement that an automatic escape and avoidance response is part of fear activation
(Barlow, 2002). Various defensive reactions such as withdrawal (flight, escape, avoidance), attentive (freezing) or tonic (unresponsive) immobility, aggressive defense, and
deflection of attack (appeasement or submission) are associated with fear arousal in
all animals including humans as a means of protection against danger (Marks, 1987).
Active avoidance of fear stimuli, which has been demonstrated in numerous animal and
human aversive conditioning experiments, is known to have reinforcing effects because
it is associated with the avoidance of punishment (Gray, 1987; Seligman & Johnston,
1973). Avoidance learning, then, is resistant to extinction because it terminates exposure to punishment (the aversive stimulus) and engenders a sense of control over the
84
COGNITIVE THEORY AND RESEARCH ON ANXIETY
situation, the latter of which augments fear reduction (for review and discussion, see
Mineka, 1979, 2004). It is not surprising that escape and avoidance responding has
played a prominent role in learning theories of fear acquisition and persistence (for further discussion, see Barlow, 2002; Craske, 2003; Öhman & Mineka, 2001; LeDoux,
1996; Marks, 1987).
Phenomenological studies of the anxiety disorders have found that some form of
immediate escape and avoidance is evident in most anxiety states. Escape and avoidance
is more prevalent in high levels of state and trait anxiety (Genest, Bowen, Dudley, &
Keegan, 1990). Most individuals with panic disorder (i.e., 90%) evidence at least mild to
moderate levels of agoraphobic avoidance (Brown & Barlow, 2002; Craske & Barlow,
1988). In social phobia individuals are more likely to engage in subtle avoidance behaviors like not giving eye contact or looking away while in social evaluative situations
(Beidel et al., 1985; Bögels & Mansell, 2004; Wells et al., 1995), whereas emotional
numbing, avoidance of trauma-related cues, or foreshortened future are active and passive avoidance responses in PTSD that reflect attempts to reduce the aversiveness of
reexperiencing the trauma (e.g., Feeny & Foa, 2006; Wilson, 2004). Between 75 and
91% of individuals with OCD have both obsessions and compulsions, the latter being
an active avoidance or escape response (Akhtar, Wig, Varma, Peershad, & Verma, 1975;
Foa & Kozak, 1995). For the vast majority of anxious patients, behavioral avoidance
plays an important role in their daily experience of this negative emotional state.
Cognitive Avoidance: An Automatic Defensive Reaction
Various cognitive processes have been identified as part of the automatic avoidance
response to threat. Attentional shift away from threat stimuli, distraction, thought suppression, and the initiation of worry are all protective cognitive processes that are aimed
at terminating or preventing exposure to threat (Craske, 2003). Ironically, these immediate responses may actually increase accessibility to the very schemas that represent
threat (Wells & Matthews, 2006). Moreover, all of these processes involve a mix of
automatic and more conscious, effortful processing. In this section we consider evidence
for an automatic cognitive avoidance, whereas the more elaborative aspects of distraction, worry, and thought suppression will be discussed as deliberate avoidant coping
strategies under Hypothesis 10.
An automatic avoidance of threat has been more consistently demonstrated in specific and social phobias than in GAD and the other anxiety disorders (see reviews by
Bögels & Mansell, 2004; Mogg & Bradley, 2004; e.g., experiment by Mogg, Bradley,
Miles, & Dixon, 2004). As a result it is still unknown whether an automatic attentional
avoidance of threat is a universal feature of all high anxiety states.
If a delayed automatic attentional avoidance of threat does emerge more consistently across the anxiety disorders, then this process could be a key element in triggering
the more conscious, strategic cognitive avoidance responses like distraction, thought
suppression, and worry (see also Mathews & Mackintosh, 1998, for similar view).
Borkovec and colleagues present compelling evidence that worry functions as a cognitive avoidance reaction to threatening information (Borkovec, 1994; Borkovec, Alcaine,
& Behar, 2004; see also Mathews, 1990) that is instigated by the automatic attentional
biases for threat. Although worry is predominantly a conscious effortful coping strat-
Empirical Status of the Cognitive Model
85
egy with an avoidant function, the initiation of the worry process may be a product of
automatic vigilance for threat.
Automatic Safety Seeking
Safety-seeking behavior is an important class of escape and avoidance behavior that
is evident in the persistence of agoraphobia (Rachman, 1984a), panic disorder (D. M.
Clark, 1997; Salkovskis, 1996a), social phobia (Rapee & Heimberg, 1997; Wells &
Clark, 1997), and PTSD (Ehlers & Clark, 2000). Various studies have shown that
increased use of safety-seeking behaviors is related to the persistence of anxiety and
avoidance (e.g., Dunmore et al., 1999; Dunmore, Clark, & Ehlers, 2001; Salkovskis et
al., 1999; Sloan & Telch, 2002; Wells et al., 1995). White and Barlow (2002) reported
that 74% of their patients with panic disorder with agoraphobia engaged in one or more
safety behaviors such as carrying a medication bottle, food/drink, bags, bracelets, or
other objects. In another study individuals with social phobia exhibited more safety
behavior that was associated with increased anxiety and that mediated actual deficits in
social performance (Stangier, Heidenreich, & Schermelleh-Engel, 2006).
Cognitive models of specific anxiety disorders and the few studies that have been
conducted on safety seeking suggest that this form of response may be important in the
pathogenesis of anxiety. However, this research is of limited relevance to Hypothesis 7
because it focuses on safety seeking as a deliberate avoidant coping strategy. Whether
there are more immediate automatic aspects of safety seeking that would make it part
of the immediate defense response is unknown at this time.
Summary
There is overwhelming clinical and laboratory evidence for a fairly automatic escape
and avoidance response in high anxiety, and this responding is part of a characteristic
automatic defensive pattern aimed at protecting the organism against threat and danger.
What is less well known is whether the elimination of escape and avoidance responses
is necessary for the successful treatment of anxiety states. Much less is known about
the more automatic features of cognitive avoidance and safety-seeking behaviors. The
research that has been published has examined these topics in terms of conscious deliberate coping strategies aimed at the reduction of anxiety. Thus more research is needed
that directly compares the automatic defensive response of high and low anxious individuals in terms of its immediate impact on anxiety level and perceived effectiveness as
a direct test of Hypothesis 7. Until this research has been conducted, the empirical status
of the cognitive and safety-seeking aspects of Hypothesis 7 is unknown.
Clinician Guideline 3.7
Relatively automatic and idiosyncratic cognitive, behavioral, and safety-seeking defensive
responses must be identified and targeted for change. A broad perspective on avoidance, one
that recognizes its cognitive and safety-seeking characteristics as part of an automatic rapid
response system to threat, is essential.
86
COGNITIVE THEORY AND RESEARCH ON ANXIETY
SECONDARY ELABORATIVE REAPPRAISAL:
THE STATE OF ANXIOUSNESS
Hypothesis 8. Facilitated Threat Elaboration
A selective threat bias will be evident in explicit and elaborated cognitive processes such that
in anxiety memory retrieval, outcome expectancies, and inferences to ambiguous stimuli will
show a preponderance of threat-related themes relative to nonanxious individuals.
As discussed in Chapter 2, the cognitive model of anxiety postulates that a secondary, compensatory stage of information processing occurs in response to threat mode
activation (i.e., the immediate fear response). Whereas the earliest moments of anxiety
are dominated by automatic processes that characterize primal threat mode activation,
the later secondary phase primarily involves deliberate and effortful processing that
reflects a conscious strategic approach to anxiety reduction.
The secondary elaborative phase plays a primary role in the persistence of anxiety.
In fact, most cognitive-behavioral interventions of anxiety focus on change at this elaborative phase. The modification of effortful cognitive processing can lead to a significant
reduction even in the more automatic aspects of fear activation. In his review Mansell
(2000) presented clinical and experimental evidence that conscious interpretations can
have a significant positive or negative impact on the automatic processes involved in
anxiety. Psychological intervention that effectively reduces anxious symptoms has been
shown to also lessen automatic attentional bias for threat (see MacLeod, Campbell,
Rutherford, & Wilson, 2004). Nevertheless, we consider conscious effortful information processing that involves making judgments, generating expectancies, evaluating or
appraising information, reasoning and decision making, and explicit memory retrieval
important aspects of the threat-biased cognitive architecture of anxiety. As evident from
the review below, there has been much debate in the research literature on the role of
elaborative, strategic processing in anxiety.
Threat-Biased Interpretations
A variety of experimental tasks have been employed to determine if anxious individuals
exhibit a greater tendency to make biased threat-related judgments than nonanxious
individuals. In some studies threat and nonthreat words were presented but evidence for
a clear preference for threat was mixed (e.g., Gotlib et al., 2004; Greenberg & Alloy,
1989). More consistent findings emerged from emotional priming experiments in which
participants are shown positive and negative trait adjectives preceded by a positive or
negative sentence prime. In these studies GAD and panic patients exhibited a preferential response to primed threat stimuli (e.g., D. M. Clark et al., 1988; Dalgleish, Cameron, Power, & Bond, 1995).
Biased judgment is more accurately investigated with experimental paradigms that
present threatening and nonthreatening ambiguous stimuli, with the prediction that
anxious individuals will endorse the more threatening interpretation. Ambiguous tasks
are more sensitive to evaluation biases because they allow for the possibility of generating alternative interpretations that vary in their aversiveness (MacLeod, 1999). One
experimental paradigm used to investigate interpretation bias involves auditory presentation of homophones, which are words with identical pronunciation but distinct spell-
Empirical Status of the Cognitive Model
87
ing, and threatening or nonthreatening meaning (e.g., die/dye; weak/week; flu/flew).
Individuals are asked to write down the word they heard presented. In an early study
Mathews, Richards, and Eysenck (1989) found that anxious patients generated significantly more threatening spellings than nonanxious patients. This finding has been replicated in other studies (e.g., Mogg, Bradley, Miller, et al., 1994, Experiments 2 and 3).
One could argue that the presentation of ambiguous sentences and other forms
of text comprehension might provide a more accurate representation of the complex
concerns we find in the anxiety disorders than single word stimuli. In these studies
anxiety disorder patients are more likely to generate or endorse threatening rather than
nonthreatening interpretations of the sentences (e.g., Amir, Foa, & Coles, 1998b; D. M.
Clark et al., 1997; Eysenck, Mogg, May, Richards, & Mathews, 1991; Harvey et al.,
1993; Stopa & Clark, 2000; Voncken, Bögels, & de Vries, 2003). On the other hand,
Constans, Penn, Ilen, and Hope (1999) found that non-socially anxious individuals had
a positive interpretation bias for ambiguous social information whereas socially anxious
individuals were more even-handed in their interpretations (see also Hirsch & Mathews,
1997). Brendle and Wenzel (2004) found that socially anxious students had particularly
pronounced negative interpretation bias to self-relevant positive unambiguous passages
and reduced positive interpretation of the same passages after 48 hours. Thus it may be
that both enhanced threat interpretation and reduced postivity bias operate differently,
especially in social phobia, but both are important in characterizing the interpretation
bias in anxiety.
One problem with homophones and ambiguous (or unambiguous) passages is that
the threatening productions of the anxious may reflect a response bias (i.e., tendency to
emit a particular response) rather than an interpretation bias (i.e., tendency to encode
or interpret stimuli in a certain threatening manner; see MacLeod, 1999). Macleod and
Cohen (1993) used a text comprehension task to show that only the high trait-anxious
students had quicker comprehension latency for ambiguous sentences that were followed by a threatening continuation sentence. This priming effect indicates that the
high but not the low trait-anxious students were more inclined to impose a threatening
meaning on the ambiguous sentences. A more recent study of homograph pairs (i.e.,
a word with two different meanings; e.g., bank could mean a financial institution or
side of a river) suggests that when threat meanings are primed in generalized social
phobia, this activated interpretative bias may persist longer than it does in nonsocially
anxious individuals (Amir et al., 2005). Furthermore, recent studies employing interpretative bias training suggest a possible causal relation between threat interpretations
and anxiety. Nonanxious individuals trained to make negative or threat interpretations
to ambiguous sentences experienced subsequent increases in state anxiety or anxiety
reactivity (Mathews & Mackintosh, 2000; Salemink, van den Hout, & Kindt, 2007a;
Wilson, MacLeod, Mathews, & Rutherford, 2006). The training effect, however, may
be more pronounced for positive interpretations (e.g., Mathews, Ridgeway, Cook, &
Yiend, 2007; Salemink et al., 2007a), with some studies even finding weak or insignificant effects of negative interpretative training on anxiety levels (Salemink, van den
Hout, & Kindt, 2007b).
In summary there is considerable evidence that the anxiety disorders are characterized by a conscious, strategic interpretation bias for threat that is particularly evident
when processing ambiguous information that is relevant to the specific anxiety concerns
of the individual. The fact that this effect has been found in priming studies indicates
88
COGNITIVE THEORY AND RESEARCH ON ANXIETY
that it can not simply be dismissed as response bias. Interpretative biases have been
demonstrated in panic disorder for body sensation information and in social phobia for
ambiguous social scenarios (see Hirsch & Clark, 2004). In addition, the interpretative
bias training studies provide evidence of a possible causal role in anxiety (see also Chapter 4). Although much remains to be understood about the specificity of the interpretative bias, we believe the findings are sufficiently well advanced to conclude that it plays
a contributory role in anxiety and so warrants a “strongly supported” designation.
Threat-Related Expectancies
If anxiety is characterized by a threat bias in elaborative processing, then anxious individuals should be more likely to hold heightened expectations for future threat or danger
that are relevant to their anxious concerns. MacLeod and Byrne (1996) reported that
anxious students anticipated significantly more negative personal future experiences
than nonanxious controls. In a 6-month follow-up of New York City workers after the
9/11 terrorist attacks, individuals who reported more PTSD symptoms also appraised
the threat of future terrorist attacks more likely (Piotrkowiski & Brannen, 2002).
Research on covariation bias indicates that heightened expectations of negative
experiences can bias perceptions of environmental contingencies (MacLeod, 1999). In
this experimental paradigm, individuals are presented fear-relevant or neutral slides that
are randomly associated with a mild shock (aversive response), a tone (neutral response),
or nothing. Participants are asked to pay attention to the stimulus– response associations
and determine whether or not there was a particular relationship between type of stimulus and response. Tomarken, Mineka, and Cook (1989) found that high fearful women
consistently overestimated the percentage of times that the fear slides were associated
with an electric shock, which reflects a processing bias for threat. This overestimation
of threat as indicated by exaggerated judgments of fear stimuli and shock associations
was replicated in spider-phobic individuals (de Jong et al., 1995), although prior fear
may have a greater effect on future covariation expectancies rather than post hoc estimates of past covariation (de Jong & Merckelbach, 2000). Covariation bias for threat
has also been demonstrated in panic-prone individuals exposed to slides of emergency
situations (Pauli, Montoya, & Martz, 1996) and, more recently, in generalized social
phobia when estimating the contingency between negative outcomes and ambiguous
social events (Hermann, Ofer, & Flor, 2004; see Garner, Mogg, & Bradley, 2006, for
contrary results). Although it is unclear whether the covariation bias is as prominent in
the anxiety disorders as it is in specific phobic states, it is evident that negative expectancies can bias judgments of contingencies that characterize anxiety-relevant situations.
Explicit Memory Bias
Information-processing research has also investigated whether anxiety is characterized
by a biased recall of threat-congruent information. If threat-relevant schemas are activated in anxiety, one would expect increased access to schema-congruent memories.
However, evidence that anxious individuals exhibit a mnemonic advantage for threatrelevant information has not been compelling (Mathews & MacLeod, 1994; MacLeod,
1999). Williams et al. (1997) concluded that biased implicit memory for threat is more
Empirical Status of the Cognitive Model
89
often found in anxiety, whereas a negative bias in explicit memory is more likely found
in depression. In addition, MacLeod (1999) concluded that anxiety vulnerability is
characterized by implicit but not explicit memory bias for threat.
Presence of an explicit memory bias for threat is indicative of bias at the strategic, elaborative phase of information processing. Contrary to earlier assertions, Coles
and Heimberg (2002) concluded in their review that explicit memory biases for threatrelevant information is evident in panic disorder and, to a lesser extent, in PTSD and
OCD. However, explicit memory bias is less apparent in social phobia and GAD.
The self-referent encoding task (SRET) has been used most often to assess explicit
memory bias in anxiety and depression. Individuals are shown a list of positive, negative
(or threatening), and neutral self-relevant words and asked to indicate which words are
self-descriptive. After the endorsement task, individuals are given an incidental recall
exercise in which they write down as many words as they can remember. Based on this
experimental paradigm or various modifications, a negative or threat recall bias has
been found for social phobia (Gotlib et al., 2004); panic disorder (Becker, Rinck, &
Margraf, 1994; Cloitre et al., 1994; Lim & Kim, 2005; Nunn, Stevenson, & Whalan,
1984); PTSD (Vrana, Roodman, & Beckham, 1995); and GAD or high trait anxiety
(Mogg & Mathews, 1990). However, other studies have failed to find a negative cued or
free recall (or recognition) bias for GAD or high trait anxiety (Bradley, Mogg, & Williams, 1995; MacLeod & McLaughlin, 1995; Mathews, Mogg, et al., 1989; Mogg et al.,
1987, 1989; Richards & French, 1991); social phobia (Cloitre, Cancienne, Heimberg,
Holt, & Liebowitz, 1995; Lundh & Öst, 1997; Rapee et al., 1994, Experiments 1 and
2; Rinck & Becker, 2005); OCD (Foa, Amir, Gershuny, et al., 1997); and even panic
disorder (Baños et al., 2001).
Coles and Heimberg (2002) noted that explicit memory bias for threat was more
apparent when conceptual or “deep” processing of information was required at the
encoding stage, when individuals did not have to produce the stimuli they fear at the
retrieval stage, when recall rather than recognition is tested, and when externally valid
experiences are used that relate directly to the fear concerns of the individual. To this
end, some researchers have investigated memory for threatening experiences by exposing individuals to imagined or real-life situations. Most of these studies involved socially
anxious individuals who were exposed to hypothetical or actual social encounters and
then assessed for encoding and retrieval of various elements of the experience. In the
majority of cases the high socially anxious group did not show an explicit threat recall
bias (e.g., Brendle & Wenzel, 2004; Rapee et al., 1994, Experiment 3; Stopa & Clark,
1993; Wenzel, Finstrom, Jordan, & Brendle, 2005; Wenzel & Holt, 2002). Radomsky
and Rachman (1999) found evidence for enhanced recall of prior contact with perceived
contamination objects (see also Radomsky, Rachman, & Hammond, 2001), but this
effect was not replicated in a later study of OCD patients with washing compulsions
(Ceschi, van der Linden, Dunker, Perroud, & Brédart, 2003).
A sufficient number of studies have found evidence of an explicit memory bias for
threat, especially when recall rather than recognition is assessed, to conclude that this
body of research provides a modest level of empirical support for Hypothesis 8. It would
appear that the conscious elaborative processing involved in the encoding and retrieval
of information may be biased toward threat in anxiety. However, an explicit memory
bias for threat has been most apparent in panic disorder and least evident in GAD and
90
COGNITIVE THEORY AND RESEARCH ON ANXIETY
social phobia. In fact, most studies have been unable to find evidence of an explicit
memory bias for threat in social phobia even with information-processing manipulations that map closely to real-life social experiences. Too few memory studies have been
conducted in OCD or PTSD to allow any conclusions to be drawn, although Muller
and Roberts (2005) recently concluded in their review that OCD is characterized by a
positive memory bias for threatening stimuli. Overall research on explicit memory bias
provides only modest support for Hypothesis 8.
Autobiographical Memory
If anxiety is characterized by threat-biased elaborative processing, then we would expect
anxious individuals to exhibit an elevated tendency to recall past personal experiences
of threat or danger. Selective retrieval of autobiographical memories has been demonstrated most clearly in depression where a negative mood- congruency effect has been
found across numerous studies (for review, see D. A. Clark et al., 1999; Williams et al.,
1997). In the typical autobiographical study, individuals are asked to report the first
memory that comes to mind in response to neutral or valenced cue words. The autobiographical memory task has good ecological validity because it assesses individuals’
personal memories and experiences, although biased recall could be caused by a greater
number of past threatening experiences in the lives of anxious individuals (MacLeod,
1999). Thus retrieval differences may not reflect memory differences as much as differences in life experiences.
Only a few studies have investigated autobiographical memory in anxiety. Rapee
et al. (1994, Experiment 4) failed to find any differences between socially anxious and
nonanxious groups in number of positive or negative memories recalled to social or
neutral stimulus words, although Burke and Mathews (1992) produced more positive
results indicative of an autobiographical memory bias in GAD. Mayo (1989) found that
high trait anxiety was associated with recall of fewer happy and more unhappy personal memories. Wenzel, Jackson, and Holt (2002) reported that individuals with social
phobia recalled more personal memories involving negative affect in response to social
threat cues but this effect was weak, accounting for only 10% of their social threat-cued
memories. Although only a few studies of autobiographical memory in anxiety have
been published, it may emerge that this memory bias may be specific to certain anxiety
disorders such as GAD but not to others like social phobia.
Summary
Overall there is considerable empirical support for Hypothesis 8, that anxiety is characterized by facilitation of threat at the elaborative, strategic stage of information processing. The strongest research support is from the interpretative bias research. The most
frequent finding is biased threat-related judgments in high anxiety. This is most apparent when ambiguous information is presented that is specific to the fear concerns of the
individual (e.g., body sensations for panic disorder and negative social evaluation for
social phobia). There is some indication that the interpretation bias in anxiety is persistent, focuses mainly on the severity of threat, and has a causal impact on anxiety. Questions still remain on whether the interpretation bias primarily involves the exaggeration
of threat or the diminution of a positivity bias that characterizes nonanxious states.
Empirical Status of the Cognitive Model
91
There is some evidence that a conscious strategic processing of threat is evident in
the form of heightened negative expectancies. Anxious individuals may be more likely
to expect that negative or threatening future events will happen to them, although more
research is needed to establish this finding. Experiments on the covariation bias indicate
that fear-related expectancies in phobic states can result in biased perceptions of environmental contingencies (MacLeod, 1999). Whether covariation biases also operate in
the anxiety disorders requires further research. However, at this stage there is at least
some experimental support for the view that anxiety involves a biased expectancy for
future negative or threatening personal events.
Finally, the considerable research literature on explicit memory bias in anxiety has
established that a biased retrieval of threat-relevant information is evident in panic disorder but not in social phobia or GAD. Too few memory studies have been conducted
on individuals with OCD or PTSD to allow firm conclusions. In addition anxious individuals may have a tendency to recall personally threatening memories and this could
contribute to other elaborative processes such as anxious rumination and postevent
processing (see Hirsch & Clark, 2004). However, evidence for selective autobiographical memory for threat is very tentative at the present time.
Clinician Guideline 3.8
Considerable empirical evidence supports therapeutic interventions that seek to change the
conscious strategic information processing that is the basis of an exaggerated reappraisal of
threat. Modify intentional threat evaluations, expectancies, and memory retrieval to establish a more balanced reappraisal of immediate threat that can have a positive impact on the
automatic processes of fear activation.
Hypothesis 9. Inhibited Safety Elaboration
Explicit and controlled cognitive processes in anxiety will be characterized by an inhibitory
bias for safety information relevant to selective threats such that memory retrieval, outcome
expectancies, and judgments of ambiguous stimuli will evidence fewer themes of safety in
comparison to nonanxious individuals.
If anxious individuals have a bias for consciously and effortfully processing threatrelevant information, is it not possible that these same strategic processes might be
biased against safety-related cues? Unfortunately, very little experimental research has
addressed this possibility. Even though a number of attentional deployment studies have
shown that anxious individuals exhibit attentional avoidance of threat stimuli at longer
presentation intervals (see discussion under Hypotheses 1 and 2), there is practically no
research on whether anxious persons show a more deliberate inhibition of safety information processing. Other researchers, such as D. M. Clark (1999), have emphasized
that safety behaviors play an important role in the persistence of anxiety, but they fail
to consider whether highly anxious individuals might actively inhibit the processing of
safety material.
In a series of experiments Hirsch and Mathews (1997) investigated the emotional
inferences that high and low anxious individuals made when primed with ambiguous
92
COGNITIVE THEORY AND RESEARCH ON ANXIETY
sentences after they read about and imaged being interviewed. The main difference
between groups occurred with the nonanxious group, who showed a quicker latency
to make positive inferences after a positive prime. The high anxiety group failed to
show this positivity bias in their online inferences. The authors concluded that biased
judgments in anxiety may be better characterized in terms of an absence of a protective
positive bias that characterizes healthy individuals (see also Hirsch & Mathews, 2000).
If we extend this deficit inferential processing of positive information to include safety
material, then these results might suggest that nonanxious individuals have a propensity
to elaborate safety-relevant information whereas individuals with social anxiety may
lack such a deliberate, strategic processing bias.
Self-report measures can also be used to assess whether anxious individuals are
less likely to deliberately process safety or corrective information. Researchers at the
Center for Cognitive Therapy in Philadelphia developed a 16-item questionnaire called
the Attentional Fixation Questionnaire (AFQ) to assess whether individuals with panic
disorder fixate on distressing physical symptoms and ignore corrective information during panic attacks (Beck, 1988; Wenzel, Sharp, Sokol, & Beck, 2005). A number of the
AFQ items deal with safety issues such as “I am able to focus on the facts,” “I can
distract myself,” “I can think of a variety of solutions,” or “I remember others’ advice
and apply it.” Fifty-five patients with panic disorder completed the questionnaire at four
time intervals: pretreatment, 4 weeks, 8 weeks, and termination. Patients who continued to have problems with panic attacks scored higher on the AFQ than individuals
with panic disorder who no longer had panic attacks, and treatment improvement was
associated with large pre–posttreatment differences on the ATQ. While only suggestive,
these results are consistent with Beck’s (1988) contention that during a panic attack
individuals are less able to consciously process safety or corrective information.
Summary
At this point it is unknown whether the interpretation threat bias in anxiety also affects
the processing of safety cues. We might expect that safety information would not be
encoded as deeply if the information-processing apparatus is oriented toward threat.
However, to date there is only suggestive evidence for inhibited or diminished elaborative processing of safety information in anxiety, with a present lack of critical research
on this issue.
Clinician Guideline 3.9
Treatment of anxiety might benefit from training that improves deliberate and effortful
processing of safety and corrective information during periods of anticipatory and acute
anxiety.
Hypothesis 10. Detrimental Cognitive Compensatory Strategies
In high anxiety states worry has a greater adverse effect by enhancing threat salience
whereas worry in low anxiety states is more likely to be associated with positive effects such
Empirical Status of the Cognitive Model
93
as the initiation of effective problem solving. In addition, other cognitive strategies aimed
at reducing threatening thoughts, such as thought suppression, distraction, and thought
replacement, are more likely to exhibit paradoxical effects (i.e., rebound, increased negative
affect, less perceived control) in high than low anxiety states.
Worry: A Maladaptive Coping Strategy
As a product of threat mode activation worry has a deleterious impact on the persistence of anxiety by enhancing the perceived likelihood and severity of threat as well as
one’s personal sense of vulnerability or ability to cope. Worry, then, has a dual function
both as a “downstream” consequence of automatic threat processes and a “feedback”
contributor to the persistence of anxiety. This leads to three specific predictions about
worry in the anxiety disorders:
•฀ Highly anxious individuals will have more excessive, exaggerated, and uncontrolled worry than those with low anxiety.
•฀ Worry in high anxiety will have a more negative consequence, resulting in greater
threat reappraisal and increased subjective anxiety.
•฀ The worry process in low anxiety is characterized by more adaptive and effective
problem solving, whereas worry in high anxiety is counterproductive.
EXCESSIVE, UNCONTROLLABLE WORRY
Considerable evidence indicates that worry is a prominent feature of all the anxiety
disorders and when it occurs in these clinical states, it is much more excessive, exaggerated, and uncontrollable than the worry reported by nonclinical individuals. In a recent
review of cognitive specificity of the anxiety disorders, it was concluded that pathological worry is not only evident in GAD but in other anxiety disorders as well, such as
panic disorder and OCD (Starcevic & Berle, 2006). Worry is a prominent feature of
symptom constructs considered common across the anxiety disorders such as anxious
apprehension (Barlow, 2002), negative affect (Barlow, 2000; Watson & Clark, 1984),
and trait anxiety (Spielberger, 1985). Although most studies find that worry is significantly more frequent, severe, and uncontrollable in GAD (Chelminski & Zimmerman,
2003; Dupuy et al., 2001; Hoyer, Becker, & Roth, 2001), nevertheless elevated levels are
also present in panic disorder, OCD, social phobia, PTSD, and even depression as well
as subsyndromal states of high anxiety (Chelminski & Zimmerman, 2003; Gladstone
et al., 2005; Wetherell, Roux, & Gatz, 2003). Naturally, the actual content of worry
will vary, with social phobia associated with social evaluative concerns, panic with the
occurrence of panic attacks or some dreaded physical consequence, PTSD with past
trauma or the negative impact of the disorder, and OCD with a variety of obsessional
fears. Moreover, the worry in GAD may be distinguished by concerns about minor daily
matters, remote future events, or illness/health/injury (Craske, Rapee, Jackel, & Barlow,
1989; Dugas, Freeston, et al., 1998; Hoyer et al., 2001). Overall, though, the research
clearly indicates that excessive and maladaptive worry is commonly associated with
states of high anxiety.
94
COGNITIVE THEORY AND RESEARCH ON ANXIETY
NEGATIVE EFFECTS OF PATHOLOGICAL WORRY
Anxious individuals worry in order to avoid unpleasant somatic anxiety or other negative emotions, as well as a problem-solving strategy that seeks to avoid or at least prepare for anticipated future negative events (Borkovec et al., 2004; Wells, 2004). In his
cognitive model of GAD, Wells (1999, 2004) emphasized that positive beliefs about the
perceived benefits of worry are an important factor in the persistence of worry and the
anxious state. However, worry is a problematic coping strategy that ultimately contributes to an escalation in anxiety by intensifying perceived threat. For clinically anxious
individuals, excessive worry will contribute to a reappraisal of the threat as even more
dangerous and imminent, and their coping resources as less than adequate for the anticipated event. Worry, then, causes an intensification of anxiety through its negative effect
on emotional responding, cognition, and ineffective problem solving.
Wells (1999) argued that the worry process is problematic because (1) it involves
the generation of numerous negative scenarios that cause a greater sense of threat and
personal vulnerability, (2) it heightens sensitivity to threat-related information, (3) it
increases the occurrence of unwanted intrusive thoughts, and (4) it leads to a misattribution of the cause for the nonoccurrence of a catastrophe, thereby strengthening positive
beliefs about worry (e.g., “I won’t do well on an exam unless I worry”).
There is considerable evidence that worry leads to an increase in subjective anxiety.
Both cross-sectional and longitudinal studies indicate that increased worry is associated with elevations in both anxiety and depression (Constans, 2001; Segerstrom, Tsao,
Alden, & Craske, 2000). The close association between repeated anxious thoughts or
worry and subjective negative emotion has been found in daily dairy studies (Papageorgiou & Wells, 1999) as well as in laboratory-based research in which nonclinical individuals are assigned to an instructed worry condition (e.g., Andrews & Borkovec, 1988;
Borkovec & Hu, 1990; York, Borkovec, Vasey, & Stern, 1987).
Another negative consequence of worry is an increase in unwanted negative intrusive thoughts. In a number of studies worry-prone individuals who engaged in a worry
induction condition later reported an increase in unwanted anxious and depressive
intrusive thoughts (Borkovec, Robinson, et al., 1983; York et al., 1987). Pruzinsky and
Borkovec (1990) found that self-labeled worriers had significantly more negative thought
intrusions than nonworriers even without a worry induction manipulation, and Ruscio
and Borkovec (2004) reported that GAD worriers had greater difficulty controlling negative thought intrusions after a worry induction than did non-GAD worriers, although
the negative intrusions caused by worry were short-lived. A causal relation between
worry and unwanted intrusive thoughts has also been demonstrated after exposure to
a stressful stimulus in which instructions to worry after viewing a film resulted in a
greater number of unwanted film intrusions (see Butler, Wells, & Dewick, 1995; Wells
& Papageorgiou, 1995).
PATHOLOGICAL WORRY, AVOIDANCE, AND PROBLEM SOLVING
The persistence of worry is a paradox. On the one hand, it is an aversive state associated
with elevated anxiety and distress, and yet we are drawn to it in times of anxiousness.
One explanation is that worry persists because of the nonoccurrence of that which
we dread (Borkovec, 1994; Borkovec et al., 2004). Moreover, it is maintained by the
Empirical Status of the Cognitive Model
95
belief that it helps in preparation for anticipated future negative outcomes (Borkovec
& Roemer, 1995). Wells (1994b, 1997) has argued persuasively that positive beliefs
about worry’s effectiveness in threat reduction contribute to its persistence. However,
the effectiveness of worry is immediately undermined by the fact that most of the things
that people worry about never happen (Borkovec et al., 2004). Under these conditions a
powerful negative reinforcement schedule is set in place in which positive beliefs about
the effectiveness of worry for avoiding or preventing bad events become strengthened by
the nonoccurrence of adverse events. So we worry not to gain any particular advantage
but instead to prevent or avoid anticipated adversity.
Even though worry may be a superfluous cognitive activity, its negative effect is further compounded by evidence that its very occurrence thwarts effective problem solving.
Measures of worry are negatively correlated with certain aspects of social problemsolving measures in both clinical and nonclinical samples (Dugas, Letarte, Rhéaume,
Freeston, & Ladouceur, 1995; Dugas, Merchand, & Ladouceur, 2005). Chronic worry
is unrelated to social problem-solving ability but more directly associated with lower
problem-solving confidence, less perceived control, and reduced motivation to engage
in problem solving (Davey, 1994; Davey, Hampton, Farrell, & Davidson, 1992; Dugas
et al., 1995). In sum, this research suggests that although pathological worry may not
be characterized by social problem-solving deficits, it probably interferes in the person’s ability to implement effective solutions (Davey, 1994). In contrast, worry phenomena in nonclinical populations may be associated with more effective implementation
of problem-solving responses (Davey et al., 1992; Langlois, Freeston, & Ladouceur,
2000b).
EXCESSIVE WORRY AND THE THREAT INTERPRETATION BIAS
A final negative consequence of worry is that it causes one to reappraise a fear stimulus
in a more threatening manner. In a study of self-reported worriers and nonworriers
in elementary school-aged children, Suarez and Bell-Dolan (2001) found that worriers
generated more threatening interpretations to hypothetical ambiguous and threatening
situations than children not prone to worry. Constans (2001) also found that worryproneness 6 weeks before an exam was associated with increased estimated risk of failing the exam. These findings, then, are consistent with our proposition that worry will
contribute to a reappraisal of threat as a more severe and probable occurrence.
Negative Impact of Safety Seeking
Even though various aspects of safety-seeking have been discussed previously, it can also
be viewed as a maladaptive compensatory coping strategy. More extensive reliance on
safety-seeking behavior has been linked to the persistence of anxiety and threat-related
beliefs (see section on Hypothesis 2). Furthermore, there is some evidence of a weaker
automatic processing of safety information and a later attentional avoidance of threat.
If more direct experimentation upholds the notion that automatic processing of safety
information is less efficient in high anxiety states, then this could help explain why the
anxious person has to expend more elaborative resources in the pursuit of safety.
Anxious individuals are more likely to utilize safety-seeking behaviors as a means
of coping with anxiety than nonanxious individuals (see section on Hypothesis 2). In
96
COGNITIVE THEORY AND RESEARCH ON ANXIETY
the short term safety-oriented coping may result in some immediate relief of anxiety
but in the long term it actually sustains threatening interpretations by preventing their
disconfirmation (Salkovskis, 1996b). In this way, extensive reliance on safety seeking
will contribute to the persistence of anxiety. The importance of safety seeking as a
maladaptive strategic coping response that contributes to the pathogenesis of anxiety
has been recognized as an important process in most of the specific anxiety disorders
such as GAD (Woody & Rachman, 1994), panic disorder (D. M. Clark, 1999), social
phobia (D. M. Clark & Wells, 1995), and PTSD (Ehlers & Clark, 2000). Like worry,
then, extensive use of safety seeking is a detrimental coping strategy that contributes to
the persistence of anxiety.
Thought and Emotion Suppression
The deliberate suppression of unwanted thoughts and emotions are two other coping
strategies that may contribute to the persistence of anxiety. Wegner and his colleagues
were the first to demonstrate that the deliberate suppression of even neutral cognitions,
such as the thought of a white bear, will cause a paradoxical rebound in the frequency of
the target thought once suppression efforts cease (Wegner, Schneider, Carter, & White,
1987). In the typical thought suppression experiment, individuals are randomly assigned
to one of three conditions: a short interval (e.g., 5 minutes) in which they can think anything except a target thought (suppression condition), an express condition (purposefully think the target thought), or monitor-only condition (think any thoughts including
the target thought). This is followed by a second interval of equal length in which all
participants are given an express or monitor-only condition. In both intervals participants indicate whenever the target thought intrudes into conscious awareness. Evidence
of postsuppression rebound is apparent when the suppression group reports a higher
rate of target intrusions during the subsequent express or monitor-only period than the
group that initially expressed or monitored their thoughts. The rebound phenomenon is
attributed to the lingering effects of intentional thought suppression that becomes most
apparent when mental control is relaxed (Wenzlaff & Wegner, 2000). The relevance of
this research for emotional disorders is obvious (for critical reviews, see Abramowitz,
Tolin, & Street, 2001; D. A. Clark, 2004; Purdon, 1999; Purdon & Clark, 2000; Rassin,
Merckelbach, & Muris, 2000; Wegner, 1994; Wenzlaff & Wegner, 2000). If unwanted
thoughts actually accelerate as a result of prior intentional suppression efforts, then
deliberate mental control of distressing thoughts would be a maladaptive cognitive coping strategy that contributes toward higher rates of threatening and disturbing cognition
seen in anxiety states. In this case thought suppression would be a major contributor
to the persistence of anxiety. However, two issues must be addressed. First, how often
do anxious individuals rely on deliberate thought suppression as a coping strategy? And
second, when anxious individuals suppress their unwanted threatening and worrisome
thoughts, is there a resurgence in anxious thinking and emotion?
PREVALENCE OF THOUGHT SUPPRESSION
The tendency to utilize thought suppression has been measured by self-report questionnaires like the White Bear Suppression Inventory (WBSI; Wegner & Zanakos, 1994).
The WBSI is a 15-item questionnaire that assesses individual differences in the tendency
Empirical Status of the Cognitive Model
97
to engage in deliberate mental control of unwanted thoughts. Positive correlations have
been reported between the WBSI and various self-report measures of anxiety as well
as measures of obsessionality (e.g., Rassin & Diepstraten, 2003; Wegner & Zanakos,
1994). Moreover, scores on the WBSI are significantly elevated in all the anxiety disorders but then decline in response to effective treatment (Rassin, Diepstraten, Merckelbach, & Muris, 2001). A factor analytic study of the WBSI, however, found that an
unwanted intrusive thoughts rather than thought suppression factor correlated with
anxiety and OCD symptoms (Höping & de Jong-Meyer, 2003). Nevertheless, other
clinical studies have indicated that thought suppression is evident in the anxiety disorders. Harvey and Bryant (1998a) found that survivors of motor vehicle accidents with
acute stress disorder (ASD) had higher ratings of natural thought suppression than survivors without ASD. A study of women who experienced a pregnancy loss revealed that
a tendency to engage in thought suppression predicted PTSD symptoms at 1 month and
4 months postloss (Engelhard, van den Hout, Kindt, Arntz, & Schouten, 2003). Overall
these findings indicate that thought suppression is a coping strategy that is very often
employed by those who are suffering from anxiety.
NEGATIVE EFFECTS OF THOUGHT SUPPRESSION
It appears that individuals with an anxiety disorder are as effective as nonclinical or
low anxious individuals in suppressing anxious target thoughts, at least in the short
term (Harvey & Bryant, 1999; Purdon, Rowa, & Antony, 2005; Shipherd & Beck,
1999), although there are other studies that indicate less efficient suppression by diagnostically anxious individuals (Harvey & Bryant, 1998a; Janeck & Calamari, 1999;
Tolin, Abramowitz, Przeworski, & Foa, 2002a). Moreover, the experimental evidence
is inconsistent in whether suppression of anxious thoughts such as worries, obsessional
intrusive thoughts, or trauma-related intrusions is more likely to result in postsuppression rebound. Some studies have reported rebound effects with anxious and obsessional
target thoughts (Davies & Clark, 1998a; Harvey & Bryant, 1998a, 1999; Koster, Rassin, Crombez, & Näring, 2003; Shipherd & Beck, 1999), whereas others have generally
failed to find any rebound suppression effects (Belloch, Morillo, & Giménez, 2004a;
Gaskell, Wells, & Calam, 2001; Hardy & Brewin, 2005; Janeck & Calamari, 1999;
Kelly & Kahn, 1994; Muris, Merckelbach, van den Hout, & de Jong, 1992; Purdon,
2001; Purdon & Clark, 2001; Purdon et al., 2005; Roemer & Borkovec, 1994; Rutledge,
Hollenberg, & Hancock, 1993, Experiment 1). Generally, it appears that postsuppression rebound of anxious thoughts is no more or less likely in clinically anxious samples
than in nonclinical individuals (see Shipherd & Beck, 1999, for contrary findings).
Even though an immediate postsuppression resurgence of unwanted thought intrusions has not been consistently supported, there is evidence that suppression of anxious
thoughts may have other negative effects that are important to the persistence of anxiety. First, it appears that over a longer time period, such as a 4- or 7-day interval, previous suppression of anxious targets will result in a significant resurgence of unwanted
thoughts (Geraerts, Merckelbach, Jelicic, & Smeets, 2006; Trinder & Salkovskis, 1994).
Abramowitz et al. (2001) suggested that individuals can successfully suppress unwanted
thoughts over short time periods, but as time progresses and individuals relax their
control efforts, a resurgence of target thought frequency is more likely. Second, suppression does appear to have a direct negative effect on mood, causing anxious and depres-
98
COGNITIVE THEORY AND RESEARCH ON ANXIETY
sive symptoms to intensify (Gaskell et al., 2001; Koster et al., 2003; Purdon & Clark,
2001; Roemer & Borkovec, 1994; Markowitz & Borton, 2002; Trinder & Salkovskis,
1994). Third, more recent studies have found that suppression of anxious or obsessional
intrusions can sustain or even alter one’s negative appraisal of their reoccurring target
intrusions and in this way contribute to an escalation in anxious mood (Kelly & Kahn,
1994; Purdon, 2001; Purdon et al., 2005; Tolin, Abramowitz, Hamlin, Foa, & Synodi,
2002b). Finally, it is clear that certain parameters can accelerate the negative effects of
suppression and/or reduce its immediate effectiveness such as the imposition of a cognitive load (see Wenzlaff & Wegner, 2000, for review) or presence of a dysphoric mood
state (Conway, Howell, & Giannopoulos, 1991; Howell & Conway, 1992; Wenzlaff,
Wegner, & Roper, 1988). Moreover, some researchers have suggested that individual
difference variables might influence the effects of suppression (Geraerts et al., 2006;
Renaud & McConnell, 2002). For example, highly obsessional individuals may be more
likely to experience persisting negative effects of suppression than individuals low in
obsessionality (Hardy & Brewin, 2005; Smári, Birgisdóttir, & Brynjólfsdóttir, 1995;
for contrary findings, see Rutledge, 1998; Rutledge, Hancock, & Rutledge, 1996).
The nature of intentional thought suppression and its role in psychopathology is currently the subject of intense empirical investigation. It is obvious that the process is complex and initial views that suppression causes a postsuppression rebound in unwanted
thought frequency that reinforces persistent emotional disturbance is overly simplified.
At the same time, the research is sufficiently clear that suppression of anxious thoughts,
especially worry, trauma-related intrusions, and obsessions, is not a healthy coping
strategy for reducing distressing thoughts and anxiety. For example, in one study individuals with panic disorder who experienced a 15-minute CO2 challenge were randomly
assigned to either accept or suppress any emotions or thoughts during the challenge test
(Levitt, Brown, Orsillo, & Barlow, 2004). Analyses revealed that the acceptance group
reported less subjective anxiety and less avoidance in response to the 5.5% CO2 challenge than the suppress group, although no differences were evident on subjective panic
symptoms or physiological arousal. At this point it is probably safe to conclude that the
intentional and effortful suppression of anxious thoughts is not a coping strategy that
should be encouraged in the management of anxiety. Rather, the expression and acceptance of distressing thoughts and images no doubt has therapeutic benefits that we are
only beginning to understand.
SUPPRESSION OF EMOTION
There has been increasing interest in the role that emotion regulation or stress reactivity
might play in specific types of psychopathology as well as psychological well-being more
generally (e.g., S. J. Bradley, 2000). One type of emotion regulation that is of particular
relevance to the anxiety disorders is emotion inhibition. Gross and Levenson (1997)
defined emotion inhibition as an active, effortful recruitment of inhibitory processes
that serve to suppress or prevent ongoing positive or negative emotion-expressive behavior. In their study of 180 undergraduate women shown amusing, neutral, and sad film
clips suppression of positive or negative emotion was associated with enhanced sympathetic activation of the cardiovascular system, reduced somatic reactivity, and a modest
decline in self-rated positive emotion.
Empirical Status of the Cognitive Model
99
Researchers have begun to investigate emotion inhibition and its related construct
of experiential avoidance in the anxiety disorders. The latter refers to an excessively
negative evaluation of unwanted thoughts, feelings, and sensations as well as to an
unwillingness to experience these private events, thereby resulting in deliberate efforts to
control or escape from them (Hayes, Strosahl, Wilson, et al., 2004b). In a study comparing Vietnam combat veterans with and without PTSD, those with PTSD reported more
frequent and intense withholding of positive and negative emotions and this tendency
to suppress emotions was specifically associated with PTSD symptomatology (Roemer,
Litz, Orsillo, & Wagner, 2001; see also Levitt et al., 2004, for panic disorder). Experiential avoidance is significantly correlated with a number of anxiety-relevant features
like anxiety sensitivity, fear of bodily sensations and suffocation, and trait anxiety, and
it prospectively predicted daily social anxiety and emotional distress over a 3-week
period (Kashdan, Barrios, Forsyth, & Steger, 2006). Although these findings are preliminary, it would appear that the suppression of emotion may join the suppression of
unwanted thoughts as a maladaptive coping strategy that inadvertently fuels distressing
emotional states like anxiety.
Clinician Guideline 3.10
Anxious individuals rely on certain deliberate and effortful coping strategies as an immediate compensation for their highly aversive subjective state. Unfortunately, any immediate
relief from anxiety due to worry, avoidance, safety-seeking behaviors, or cognitive/experiential suppression is temporary. Indeed, these strategies actually play a prominent role in
the longer term persistence of anxiety states. Thus effective intervention must redress the
detrimental impact that these maladaptive effortful coping strategies have on anxiety.
SUMMARY AND CONCLUSION
A review of the research literature relevant to the cognitive model of anxiety (see Figure
2.1) indicates there is mounting empirical support for the role of automatic cognitive
processes in immediate fear activation. This is most evident for Hypothesis 1, where
there is consistent experimental data indicating that fear is characterized by an automatic, preconscious attentional threat bias for moderately intense personal threat stimuli presented at very brief exposure intervals. Very little research has been conducted on
the possibility of an automatic attentional processing against safety information (i.e.,
Hypothesis 2), although there is moderate research support for an automatic threat
evaluation process in high anxiety states (i.e., Hypothesis 3).
Hypotheses 4 to 7 focus on various cognitive, behavioral, and emotional consequences elicited by immediate threat mode activation. There is considerable evidence that
anxious individuals overestimate the probability, proximity, and, to a lesser extent, the
severity of threat-relevant information (i.e., Hypothesis 4). There is consistent empirical
evidence that highly anxious individuals misinterpret their anxious symptoms in a negative or threatening manner (i.e., Hypothesis 5) and that automatic negative thoughts and
images of threat, danger, and personal vulnerability or helplessness characterize anxiety
100
COGNITIVE THEORY AND RESEARCH ON ANXIETY
states (i.e., Hypothesis 6). However, research on cognitive content-specificity was much
less consistent in demonstrating that threatening thought content is specific to anxiety.
It may be that cognitive specificity would be more apparent if researchers focused on
disorder-specific cognitions rather than on general forms of apprehensive thought.
Hypothesis 7, which proposes that an automatic defensive response is elicited
by immediate threat mode activation, has mixed support. Although there is a wellestablished behavioral literature demonstrating the prominence of behavioral escape as
an automatic defensive response in anxiety, there has been little research on an automatic cognitive avoidance and safety-seeking defensive response.
The final three hypotheses reviewed in this chapter deal with the secondary, elaborative phase of anxiety. This component of the anxiety program will be of greatest
interest to practitioners because the processes involved in the elaboration of anxiety
have a direct impact on its persistence. This is also the phase that is specifically targeted
in cognitive therapy of anxiety. Empirical support for Hypothesis 8 was strong, with
numerous studies demonstrating that anxious individuals exhibit a deliberate threat
interpretation bias for ambiguous stimuli, which is indicative of a conscious, strategic
threat-processing bias. However, it is unknown whether diminished elaborative processing of safety information occurs in anxiety (i.e., Hypothesis 9) because there is practically no research on the topic. Empirical evidence for maladaptive cognitive coping
strategies in anxiety is very strong (i.e., Hypothesis 10), with numerous studies demonstrating the detrimental effects of worry, excessive safety-seeking behavior, thought
suppression, and, more recently, experiential avoidance. This research clearly highlights
the importance of targeting these response strategies when offering cognitive therapy
for anxiety.
Our extensive review of the extant empirical research clearly supports a cognitive
basis to anxiety. Specific cognitive structures, processes, and products are critical to the
activation and persistence of anxiety. Although this research provides a basis for advocating a cognitive approach to the treatment of anxiety, it does not address the question
of etiology. In the next chapter we consider whether there might be a causal role for
cognition in the etiology of anxiety.
Chapter 4
Vulnerability to Anxiety
We walk in circles so limited by our own anxieties that we can no longer distinguish
between true and false, between the gangster’s whim and the purest ideal.
—I NGRID BERGMAN (Swedish-born actress, 1915–1982)
P
eople who have suffered years from an anxiety disorder are often perplexed about the
origins of their disorder. Clients will frequently ask “Why me?,” “How come I developed this problem with anxiety?”, “Did I inherit this condition, do I have some kind of
imbalance in my brain chemistry?,” “Did I do something to bring this on myself?” “Is
there a flaw in my personality or some weakness in my psychological makeup?” Unfortunately, clinicians facing questions about the etiology of anxiety have great difficulty
providing satisfactory answers given that our knowledge of vulnerability to anxiety is
relatively limited (McNally, 2001).
Even though research on vulnerability has lagged behind our knowledge of the psychopathology and treatment of anxiety, most would agree that susceptibility to developing an anxiety disorder varies greatly within the general population. This is well
illustrated in the following case examples. Cynthia, a 29-year-old factory worker, who
described herself as being highly anxious, worrisome, and lacking self- confidence since
early childhood, developed moderately severe doubt and checking compulsions after
leaving high school and assuming the increased responsibilities of work and living independently. Andy, a 41-year-old accountant, presented with a first onset of severe panic
disorder and agoraphobic avoidance after promotion to a highly stressful, performancedriven managerial position that led to the onset of various physical symptoms, such as
chest pressure and pain, heart palpitations, numbness, sweating, lightheadedness, and
stomach tightness. He had a comorbid health anxiety that intensified after receiving
treatment for hiatus hernia, high cholesterol, and acid reflux. Ann Marie, a 35-yearold government office worker, suffered from long-standing social phobia that remained
untreated until she experienced her first full-blown panic attack after a promotion that
caused a significant increase in her work stress. Ann Marie stated that she had always
been a generally anxious and worrisome person since high school, but currently found
social interactions the most threatening for her.
101
102
COGNITIVE THEORY AND RESEARCH ON ANXIETY
In each of these case illustrations the emergence of an anxiety disorder occurred
within the context of predisposing factors and precipitating circumstances. Frequently
individuals with anxiety disorders report a predisposition toward high anxiety, nervousness or worry, as well as precipitating events that escalate their daily stress. Since
predisposing biological or psychological characteristics and environmental factors are
both involved in the etiology of clinical anxiety, diathesis– stress models are frequently
proposed to account for individual differences in risk for anxiety (Story, Zucker, &
Craske, 2004). In many cases major life events, traumas, or ongoing adversities are
involved in anxiety; in others, the precipitants are not so dramatic, and fall within the
realm of normal life events (e.g., increased work stress, an uncertain medical test, an
embarrassing experience). These differences in clinical presentations have led researchers to search for vulnerability and risk factors that might predict whether a person
develops an anxiety disorder.
In this chapter we present the cognitive model of vulnerability to anxiety. We begin
by defining some of the key concepts employed in etiological models of disorder. This is
followed by an overview of the role that heritability, neurophysiology, personality, and
life events may play in the origins of anxiety disorders. We then present the cognitive
vulnerability model of anxiety that was first articulated in Beck et al. (1985). The chapter concludes with a discussion of the empirical support for the last two hypotheses of
the cognitive model, elevated personal vulnerability and enduring threat-related beliefs,
that pertain directly to the issue of etiology.
VULNERABILITY: DEFINITIONS AND CARDINAL FEATURES
Although often used interchangeably, the terms “vulnerability” and “risk” have very
distinct meanings (see Ingram, Miranda, & Segal, 1998; Ingram & Price, 2001). Risk
is a descriptive or statistical term referring to any variable whose association with a
disorder increases its likelihood of occurrence (e.g., gender, poverty, relationship status)
without informing about actual causal mechanisms. Vulnerability, on the other hand,
is a risk factor that has causal status with the disorder in question. Vulnerability can be
defined as an endogenous, stable characteristic that remains latent until activated by a
precipitating event. This activation can lead to the occurrence of the defining symptoms
of a disorder (Ingram & Price, 2001). Knowledge of vulnerability factors has treatment
implications because it will elucidate the actual mechanisms of etiology (Ingram et al.,
1998). However, vulnerability does not directly lead to disorder onset but instead is
mediated by the occurrence of precipitating events.
Vulnerability factors are internal, stable, and latent or unobservable until activated
by a precipitating event (Ingram et al., 1998; Ingram & Price, 2001). This private, unobservable nature of vulnerability in asymptomatic individuals has presented special challenges for researchers in search of reliable and valid methods for detecting vulnerability
(Ingram & Price, 2001). Moreover, vulnerability constructs must have high sensitivity
(i.e., must be present in disordered individuals), a moderate level of specificity (i.e., more
prevalent in the target disorder than in controls), and be distinct from the precipitating
life event (Ingram et al., 1998). In Beck’s cognitive model vulnerability constructs are
neither necessary nor sufficient but rather contributory causes of psychopathology that
may interact or combine with other etiological pathways that are present at the genetic,
Vulnerability to Anxiety
103
biological, and developmental levels (see Abramson, Alloy, & Metalsky, 1988; D. A.
Clark et al., 1999).
The cognitive model of anxiety presented in Chapter 2 (see Figure 2.1) describes
the proximal cognitive structures and processes involved in the persistence of anxiety,
whereas this chapter focuses on distal cognitive variables that are predispositions for
anxiety. These distal cognitive vulnerability factors are moderators (i.e., they affect the
direction and/or strength of association between stress and symptom onset), whereas
more proximal cognitive variables are mediators (i.e., they account for the relationship
between vulnerability, stress, and disorder onset) (see Baron & Kenny, 1986; Riskind &
Alloy, 2006). In the cognitive model multiple distal vulnerabilities are present at the biological, cognitive, and developmental levels so that some individuals may have multiple
vulnerabilities. These compound vulnerabilities might be associated with even higher
risk for disorder onset, a more severe symptom presentation, or comorbid emotional
conditions (Riskind & Alloy, 2006).
BIOLOGICAL DETERMINANTS
Individual differences in genetics, neurophysiology, and temperament will interact with
a predisposing cognitive vulnerability to heighten or reduce one’s anxiety-proneness in
response to life adversity or threat. Barlow (2002) convincingly argued for a generalized
biological vulnerability to anxiety disorders, in which heritability, a nonspecific vulnerability factor, accounts for 30–40% of variability across all anxiety disorders. This
genetic vulnerability is most likely expressed through elevations in broad personality
traits or temperaments like neuroticism, trait anxiety, or negative affectivity. Chronic
arousal, prepotent neuroanatomical structures (e.g., amygdala, locus coeruleus, BNST,
right prefrontal cortex), and neurotransmitter abnormalities in serotonin, GABA, and
CRH are other biological vulnerabilities to anxiety that have etiological significance, in
part by interacting in a synergistic fashion with cognitive vulnerability (see Chapter 1
for further discussion).
PERSONALITY VULNERABILITY
Neuroticism and Negative Affectivity
Eysenck and Eysenck (1975) described neuroticism (N) as a predisposition toward emotionality in which the highly neurotic individual is overly emotional, anxious, worrisome, moody, and has a tendency to overreact strongly to a range of stimuli. High
N and low E (extraversion) individuals—or introverted individuals—were considered
more likely to develop anxiety because they have an overreactive limbic system that
causes them to more easily acquire conditioned emotional responses to arousing stimuli.
Although there is strong empirical support for high N in the pathogenesis of anxiety
(e.g., see review by Watson & Clark, 1984), empirical evidence for other characteristics
of N, such as its neurophysiological basis, have not been as well supported (Eysenck,
1992).
Watson and Clark (1984) proposed a mood– dispositional dimension called negative affectivity (NA). NA reflects a “pervasive individual difference in negative emo-
104
COGNITIVE THEORY AND RESEARCH ON ANXIETY
tionality and self- concept” (p. 465), with high NA individuals more likely to experience
elevated levels of negative emotions including subjective feelings of nervousness, tension,
and worry, as well as a tendency to have poor self-esteem and to dwell on past mistakes, frustrations, and threats (Watson & Clark, 1984). Research within the Big Five
personality tradition have subsumed the notion of N and NA under the higher order,
superordinate personality construct of “negative emotionality” (e.g., Watson, Clark, &
Harkness, 1994).
There is a large correlational and factor analytic research showing an association
between negative emotionality and anxiety in clinical and nonclinical samples (i.e., Longley, Watson, Noyes, & Yoder, 2006). Higher emotionality is evident in all the anxiety disorders as well as in depression (e.g., Bienvenu et al., 2004; Cox, Enns, Walker,
Kjernisted, & Pidlubny, 2001; Trull & Sher, 1994; Watson, Clark, & Carey, 1988) and
it predicts future anxious symptoms (Gershuny & Sher, 1998; Levenson, Aldwin, Bossé,
& Spiro, 1988). Thus high NA or emotionality is a broad, nonspecific distal vulnerability factor for anxiety and its disorders that constitutes a temperamental characteristic of
proneness to nervousness, tension, and worry with roots in genetic and early childhood
experiences (i.e., Barlow, 2002).
Trait Anxiety
Another personality construct so closely related to negative emotionality (i.e., N or
NA) that the two are considered almost synonymous is trait anxiety (Eysenck, 1992).
Spielberger, the strongest proponent for distinguishing between state and trait anxiety,
defined state anxiety as “a transitory emotional state or condition of the human organism that is characterized by subjective, consciously perceived feelings of tension, apprehension and heightened autonomic nervous system activity. A-States vary in intensity
and fluctuate over time” (Spielberger, Gorsuch, & Lushene, 1970, p. 3).
Trait anxiety, on the other hand, is considered to be “relatively stable individual
differences in anxiety proneness” (Spielberger et al., 1970, p. 3). Individuals with high
trait anxiety are more likely to respond to situations of perceived threat with elevations
in state anxiety and evaluate a greater range of stimuli as threatening, have a lower
anxiety activation threshold, and feel more intense anxious states (Rachman, 2004;
Spielberger, 1985). Although there is substantial evidence that Spielberger’s State-Trait
Anxiety Inventory is highly relevant for stress and anxiety (Roemer, 2001), high trait
anxiety is a problematic vulnerability construct because (1) its temporal stability has not
been consistently supported, (2) its unidimensional structure has been challenged, (3)
it is too highly correlated with state anxiety, (4) it may lack specificity for anxiety, and
(5) it embodies a vague idea of vulnerability that is closely aligned with Freud’s concept
of neurotic anxiety (Eysenck, 1992; Rachman, 2004; Reiss, 1997; Roemer, 2001). For
these reasons researchers have looked elsewhere for more specific personality predictors
of anxiety disorders.
Anxiety Sensitivity
In recent years anxiety sensitivity, a fear of or sensitivity to experiencing anxiety, has
emerged as a more promising personality vulnerability construct that takes a more cog-
Vulnerability to Anxiety
105
nitive perspective with greater specificity to anxiety and its disorders. Anxiety sensitivity (AS) is the fear of anxiety-related bodily sensations based on enduring beliefs that
negative physical, social, or psychological consequences might result from these anxious
symptoms (Reiss, 1991; Reiss & McNally, 1985; Taylor, 1995a; Taylor & Cox, 1998).
For example, a person with high AS might interpret chest pain as sign of an imminent heart attack and so feel highly anxious when experiencing this bodily sensation,
whereas a person with low AS might interpret the chest pain as muscle tension due to
physical exertion and so experience no anxiety with the bodily sensation.
A propensity to feel anxious about certain bodily symptoms is present in high AS
because individuals believe anxiety and its physical symptoms can lead to serious consequences like heart attacks, mental illness, or intolerable anxiety (Reiss, 1991). Thus
AS is a personality variable that amplifies fear when anxiety sensations and behaviors
are experienced (Reiss, 1997). In this way it is thought to play both an etiological and a
maintaining role in all the anxiety disorders, but particularly panic disorder and agoraphobia (Reiss, 1991; Taylor & Cox, 1998).
Psychometric Validation
The 16-item Anxiety Sensitivity Index (ASI) is the primary measure for assessing individual differences in AS (Reiss, Peterson, Gursky, & McNally, 1986; Reiss & McNally,
1985). Despite considerable debate over its factorial structure, it now appears that the
ASI is a hierarchical multidimensional construct with two or three correlated lower
order factors (i.e., Fears of Mental Catastrophe vs. Fears of Cardiopulmonary Sensations or Physical Concerns, Mental Incapacitation, and Social Concerns about Being
Anxious) linked to a higher order general factor of AS (Mohlman & Zinbarg, 2000;
Schmidt & Joiner, 2002; Zinbarg, Barlow, & Brown, 1997). There is also controversy
over which dimensions best describe AS. Based on a 36-item ASI-R, only two correlated
factors were replicated across data sets drawn from six countries: Fear of Somatic Symptoms and Social- Cognitive Concerns (Zolensky et al., 2003).
The most recent revision of the ASI, the 18-item ASI-3, may provide the best assessment of the three AS dimensions; physical, cognitive, and social concerns (Taylor, Zvolensky, et al., 2007). The ASI-3 subscales had improved internal consistency and good
criterion-related validity, although the three subscales were very highly correlated (r’s >
.83). Nevertheless, findings across the various versions of the ASI indicate that subscales
rather than a total score should be utilized to indicate level of AS.
The ASI measures have good internal consistency, test–retest reliability, and strong
convergent validity with other measures of anxiety (Mohlman & Zinbarg, 2000; Reiss
et al., 1986; Taylor & Cox, 1998; Zvolensky et al., 2003). Moreover, the AS lower order
dimensions are generally consistent across various countries (Bernstein et al., 2006; Zvolensky et al., 2003), although there is some evidence that high AS scores may decrease
over time even in the absence of a specific intervention (Gardenswartz & Craske, 2001;
Maltby, 2001; Maltby, Mayers, Allen, & Tolin, 2005). There has been considerable
debate on whether AS is distinct from trait anxiety (for discussion, see Lilienfeld, 1996;
Lilienfeld, Jacob, & Turner, 1989; McNally, 1994). The current view is that AS is a distinct lower order construct hierarchically linked to the broader personality disposition
of trait anxiety (Reiss, 1997; Taylor, 1995a).
106
COGNITIVE THEORY AND RESEARCH ON ANXIETY
Experimental Validation
If AS amplifies fear reactions, then high AS should lead to more intense anxiety in
response to a wider range of stimuli (Reiss & McNally, 1985; see Taylor, 2000). This
should be particularly evident in biological challenges that provoke panic attacks under
controlled laboratory conditions or other experimental manipulations that elicit the
physical symptoms of anxiety (McNally, 1996). In fact, there is now considerable empirical evidence that baseline AS predicts postchallenge anxiety symptoms and panic attacks
in people with or without diagnosable panic disorder (for reviews, see McNally, 2002;
Zvolensky, Schmidt, Bernstein, & Keough, 2006). High AS predicts fearful response
and panic symptoms to carbon dioxide (CO2) inhalation (e.g., Rapee, Brown, Antony, &
Barlow, 1992; Rassovsky, Kushner, Schwarze, & Wangensteen, 2000; Schmidt & Mallott, 2006), hyperventilation (Carter, Suchday, & Gore, 2001; Holloway & McNally,
1987; McNally & Eke, 1996; Rapee & Medoro, 1994), and caffeine ingestion (Telch,
Silverman, & Schmidt, 1996). Although ASI Physical Concerns may be the only AS
dimension that predicts fear response to a physical challenge (Brown, Smits, Powers,
& Telch, 2003; Carter et al., 2001; Zvolensky, Feldner, Eifert, & Stewart, 2001), these
experimental findings support the predictive validity of the ASI and its special relevance
to panic-spectrum psychopathology (Zvolensky et al., 2006).
Diagnostic Specificity
If AS is a specific cognitive-personality vulnerability factor for anxiety, then it should
be significantly more elevated in anxiety, especially panic disorder, than in other clinical and nonclinical samples (McNally, 1994, 1996). Individuals with panic disorder or
agoraphobia score on average two standard deviations above the normative mean on
the ASI (McNally, 1994, 1996; Reiss, 1991; Taylor, 1995a, 2000) and anxiety disorder
samples (except simple phobias) score significantly higher than depression or nonclinical
comparisons (Taylor & Cox, 1998; Taylor, Koch, & McNally, 1992). Within the anxiety disorders, persons with panic disorder and agoraphobia score significantly higher
than the other anxiety disorders, with PTSD, GAD, OCD, and social phobia scoring significantly higher than nonclinical comparison groups (Deacon & Abramowitz,
2006a; Taylor, Koch, & McNally, 1992a). At the symptomatic level ASI has a specific
association with self-report of panic attacks in nonclinical child and adult populations
(e.g., Calamari et al., 2001; Cox, Endler, Norton, & Swinson, 1991; Longley et al.,
2006), although some studies have found AS relates to depressive symptoms as well
(Reardon & Williams, 2007).
The ASI subscales appear to have differential specificity for anxiety and panic.
ASI Physical Concerns is the only dimension specific to panic disorder whereas Social
Concerns may be more relevant to social phobia (e.g., Deacon & Abramowitz, 2006a;
Zinbarg et al., 1997) and Cognitive Dyscontrol may be related to depression (Cox et
al., 2001; Rector, Szacun-Shimizu, & Leybman, 2007). However, caution must be exercised when using the ASI to screen for anxiety or panic. Hoyer and colleagues examined
the predictive accuracy of the ASI, BAI, and several other anxiety measures in a large
epidemiological sample of 1,877 young women in Dresden, Germany (Hoyer, Becker,
Neumer, Soeder, & Margraf, 2002). None of the measures alone were able to accu-
Vulnerability to Anxiety
107
rately screen for anxiety disorders, although better predictive accuracy occurred when
a specific anxiety disorder was targeted by more specific symptom questionnaires (e.g.,
screening for agoraphobia with the Mobility Inventory). Clearly, then, it would be incorrect to assume the presence or absence of panic solely on the basis of an individual’s ASI
score.
Prospective Studies
The best empirical evidence that AS is a cognitive personality vulnerability factor for
panic disorder comes from longitudinal studies. Maller and Reiss (1992) reported that
ASI scores predicted frequency and intensity of panic attacks 3 years later. In two separate samples of U.S. Air Force cadets assessed before and after a stressful 5 weeks of basic
cadet training, the ASI predicted spontaneous panic attacks that occurred in 6% of the
cadets during the 5-week period (Schmidt, Lerew, & Jackson, 1997, 1999). Additional
analyses revealed that AS uniquely predicted changes in anxious symptoms (i.e., BAI
scores) when controlling for the close association between anxiety and depression. Unexpectedly, analysis of the ASI subfactors revealed that it was ASI Mental rather than Physical Concerns that predicted the spontaneous panic attacks and changes in BAI scores.
In a 4-year community-based longitudinal study, adolescents classified as stable
high or escalating ASI scorers were significantly more likely to experience a panic attack
than low stable scorers (Weens, Hayward, Killen, & Taylor, 2002). However, there was
little evidence that experiencing panic led to subsequent increases in AS (see Schmidt,
Lerew, & Joiner, 2000, for contrary findings). Plehn and Peterson (2002) conducted
an 11-year mailed follow-up survey with first year undergraduates initially assessed for
AS and trait anxiety. After controlling for history of panic symptoms, only Time 1 ASI
was a significant predictor of panic symptoms and panic attacks over an 11-year time
interval. Surprisingly, trait anxiety, not AS, was the only significant predictor of panic
disorder. In a retrospective cross-sectional study ASI Physical Concerns and exposure
to aversive life circumstances predicted panic attacks and agoraphobic avoidance in
the past week (Zvolensky, Kotov, Antipova, & Schmidt, 2005). Together these findings
indicate that high AS constitutes a significant cognitive-personality predisposition for
panic attacks. However, it is unclear which of the ASI subfactors is the most potent
predictor of panic and whether having panic causes a “scarring effect” on AS (i.e., leads
to subsequent increase in AS). McNally (2002) also reminds us that the amount of variance accounted for by AS is modest, suggesting that other factors are clearly important
in the etiological of panic.
Treatment Effects
There is considerable evidence that AS is responsive to interventions (for reviews, see
McNally, 2002; Zvolensky et al., 2006). For example, a primary preventive program
that targeted AS produced significant reductions in AS that translated into lower subjective fear response to a biological challenge and a significant decrease in Axis I psychopathology over a 2-year follow-up period (Schmidt, Eggleston, et al., 2007). Thus
targeting AS in cognitive therapy should produce immediate and long-term benefits in
anxiety reduction.
108
COGNITIVE THEORY AND RESEARCH ON ANXIETY
Anxiety Sensitivity and the Cognitive Model
Empirical evidence that AS is a specific predisposing factor for anxiety, especially panic,
fits with the cognitive vulnerability model of anxiety. AS is a cognitive construct that
describes individual differences in the propensity to misinterpret the bodily sensations
of anxiety in a threatening manner. It is a specific cognitive vulnerability construct that
may have relevance beyond panic to the extent that negative interpretation of subjective
anxiety and its symptoms is a consequence of automatic threat mode activation (see
Chapter 2). In Chapter 3 we discussed empirical evidence that emotional reasoning or
a tendency to interpret anxious symptoms in a negative or threatening manner is an
important cognitive phenomenon in anxiety. We expect that individuals high in AS will
more likely engage in emotional reasoning and other forms of biased interpretation of
their anxious symptoms than individuals low in AS.
Based in part on correlational analyses between the ASI and the Fear Survey Schedule (see Taylor, 1995a), Rachman (2004) concluded that AS along with illness/injury
sensitivity and fear of negative evaluation are distinct lower order traits that are nested
hierarchically in the broader construct of trait anxiety. All three of these constructs
are cognitive in nature since they focus on a tendency to misinterpret physical or social
information in a negative or threatening manner. They describe specific cognitivepersonality vulnerabilities for panic and social- evaluative anxiety states. And yet, even
though there is strong empirical support that AS is a vulnerability factor in anxiety, its
ability to account for only modest variance indicates that other cognitive-personality
factors must be involved in the pathogenesis of anxiety disorder.
Clinician Guideline 4.1
Include the ASI or ASI-3 in the pretreatment assessment battery to evaluate the client’s propensity to misinterpret physical, cognitive, and social symptoms in an anxious or fearful
manner.
Diminished Personal Control
It has been suggested that the greatest human fear is of losing control, leading many
researchers to consider impaired control a key feature of stress, anxiety, depression, and
other aspects of psychological distress (Mineka & Kihlstrom, 1978; Shapiro, Schwartz,
& Astin, 1996). In his account of the origins of anxious apprehension, Barlow (2002)
posited that a generalized psychological vulnerability interacts with a generalized biological vulnerability and particular learning experiences in the development of specific
anxiety disorders. Psychological vulnerability was defined as “a chronic inability to
cope with unpredictable uncontrollable negative events, and this sense of uncontrollability is associated with negatively valenced emotional responding” (Barlow, 2002,
p. 254). Earlier Chorpita and Barlow (1998) defined control as “the ability to personally
influence events and outcomes in one’s environment, principally those related to positive
or negative reinforcement” (p. 5).
In anxiety uncertainty that one possesses the required level of control over an anticipated aversive outcome is an enduring characteristic (Alloy et al., 1990). This dimin-
Vulnerability to Anxiety
109
ished sense of personal control is an individual difference variable that may be acquired
through childhood experiences of stifled independence, limited exploration, and high
parental protection. As a result of repeated experiences of uncontrollable or unpredictable events throughout early and middle childhood, the individual develops low perceived control over life circumstances and perhaps increased neurobiological activity in
the behavioral inhibition system (Barlow, 2002; Chorpita & Barlow, 1998). According
to Barlow, these beliefs of low personal control constitute a personality diathesis that
interacts with negative or aversive life events to trigger anxiety or depression.
It has been long recognized that a decrease in perceived control is associated with
anxiety and that lower control over a threatening event can increase estimates of the
probability of danger and personal vulnerability (Chorpita & Barlow, 1998). Beck et
al. (1985, 2005) recognized that fear of losing control is a prominent cognitive feature
found in many anxiety states. Barlow and colleagues (Barlow, 2002; Chorpita & Barlow, 1998) note that the perception that threatening events occur in an unexpected,
unpredictable fashion is part of a diminished sense of personal control over aversive
events. However, there is a lack of direct support for a specific association between
chronic diminished control and anxiety (see Barlow, 2002; Chorpita & Barlow, 1998).
In fact there has been a long research tradition in locus of control, learned helplessness,
life event appraisals, and attributional style that recognizes a role for perceived control
in depression (e.g., Abramson, Metalsky, & Alloy, 1989; Alloy, Abramson, Safford, &
Gibb, 2006; Hammen, 1988). Alloy et al. (1990), for example, stated that a generalized
tendency to perceive negative events as uncontrollable is a distal contributory cause of
depression.
Alloy and colleagues further proposed a helplessness– hopelessness theory that identifies certain key cognitive processes that underlie the high comorbidity between anxiety
and depression (Alloy et al., 1990). According to the theory, anxiety is experienced when
individuals expect to be helpless in controlling important future events but are uncertain about their helplessness, whereas this anxiety turns to hopelessness and depression
when the future negative outcomes become certain. Unfortunately, research on the role
of a cognitive style of diminished control for negative outcomes in anxiety, and its probable connection to depression, is limited (Chorpita & Barlow, 1998). This state of affairs
is partly due to the lack of sensitive measures of perceived uncontrollability of threat.
To rectify this situation, the 30-item Anxiety Control Questionnaire (ACQ) was developed to assess perceived control over anxiety-related symptoms, emotional reactions,
and external problems and threats (Rapee, Craske, Brown, & Barlow, 1996). The ACQ
has good internal consistency, 1-month test–retest reliability, and moderate correlations
with anxiety and depression symptom measures (see also Zebb & Moore, 1999).
There is fairly consistent empirical evidence of a close association between anxiety
and diminished sense of control over outcomes. In a panic disorder study agoraphobic
avoidance was greatest in those who had high anxiety sensitivity and low ACQ perception of control (White, Brown, Somers, & Barlow, 2006). Likewise Hofman (2005)
found that anxiety in social phobia persists because individuals have low perceived
control over emotions and bodily sensations when exposed to social threat (see also
McLaren & Crowe, 2003; Rapee, 1997, for similar findings).
Despite evidence of an association between diminished sense of control over potentially threatening outcomes and anxiety, there is a significant body of research from the
110
COGNITIVE THEORY AND RESEARCH ON ANXIETY
attributional style literature showing that reduced perceived control over past negative
events may have an even stronger relation to depression than to anxiety. A negative or
pessimistic attributional style refers to the belief that the cause of past loss and failure can be attributed to internal, global, and stable or enduring personal deficiencies
(Abramson, Seligman, & Teasdale, 1978). A negative attributional style can be viewed
as a diminished sense of past control. There is now considerable evidence that negative
attributional style is a cognitive-personality vulnerability for depression (for reviews,
see Alloy et al., 2006; Sweeney, Anderson, & Bailey, 1986; e.g., Hankin, Abramson,
& Siler, 2001; Metalsky, Halberstadt, & Abramson, 1987). However, studies that have
examined the specificity of the negative attributional style reveal that it is also apparent
in anxiety, though to a lesser degree, (e.g., Heimberg et al., 1989; Johnson & Miller,
1990; Luten, Ralph, & Mineka, 1997).
Perceived reduction in control over potentially threatening outcomes appears to be
an important factor in the anxiety disorders, especially if there is elevated uncertainty
about the threat (Alloy et al., 1990; Moulding & Kyrios, 2006). However, the necessary
longitudinal research has not been conducted to determine whether there is an enduring
belief in diminished personal control over threat that is a distal contributory factor to
anxiety. Nevertheless, there is sufficient evidence to conclude that low perceived control
is a contributor to anxiety, although it is probably a nonspecific cognitive-personality
factor found in both depression and anxiety.
Clinician Guideline 4.2
Include assessment of perceived control over threat in the case formulation. Two aspects
of control are important to assess in anxiety: (1) clients’ perceived control over emotional
responses, especially symptoms of anxiety; and (2) clients’ evaluations of their ability to
manage anticipated threats related to their primary threat concerns. The ACQ can be helpful when assessing perceived control of anxiety.
LIFE EVENT PRECIPITANTS OF ANXIETY
Diathesis– stress models have been proposed for anxiety that explain disorder onset
in terms of an interaction between negative life events and a preexisting vulnerability
diathesis (e.g., Barlow, 2002; Chorpita & Barlow, 1998). A life event, situation, or circumstance that is evaluated as a potential threat to one’s survival or vital interests can
activate an underlying vulnerability that will lead to a state of anxiety. This underlying
diathesis can involve personality predispositions like high negative emotionality, trait
anxiety, anxiety sensitivity, and a chronic sense of diminished control, as well as more
specific cognitive vulnerabilities such as hypervalent threat schemas and heightened
sense of personal weakness and ineffectiveness (see discussion below).
There is evidence that an excess of negative life events is associated with the anxiety disorders. In a large population-based twin study, the occurrence of high-threat life
events was associated with a significant increase in risk of developing a pure generalized anxiety episode (e.g., Kendler, Hettema, Butera, Gardner, & Prescott, 2003). In
a retrospective study of life adversity and onset of psychiatric disorders in over 1,800
Vulnerability to Anxiety
111
community-based young adults, individuals who averaged over six major life events or
potentially traumatic experiences and an accumulating exposure to lifetime adversity
had increased risk of depressive or anxious episodes (Turner & Lloyd, 2004). Stressful
or adverse life experiences have been shown to frequently precede and/or exacerbate
the onset of OCD (i.e., Cromer, Schmidt, & Murphy, 2007), social phobia, panic disorder, GAD, and, of course, PTSD (for reviews, see Clark, 2004; Craske, 2003; Ledley,
Fresco, & Heimberg, 2006; Taylor, 2000, 2006). However, one must keep in mind that
many individuals develop an anxiety disorder without experiencing a major negative life
event, and most people who experience life adversities never develop an anxiety disorder
(McNally, Malcarne, & Hansdottir, 2001).
Although there is consistent evidence that life events play an etiological role in
anxiety, it is also apparent that their contribution may be less prominent in anxiety
than they are in depression. For instance, Kendler, Myers, and Prescott (2002) did not
find evidence to support a diathesis– stress model for the acquisition of phobias (see also
Brown, Harris, & Eales, 1996). Thus threatening life events and other experiences of
life trauma and adversity are significant contributors in the pathogenesis of anxiety, but
much remains to be learned about the exact nature of these proximal contributors and
how they interact with the cognitive-personality vulnerability factors for anxiety.
Clinician Guideline 4.3
Given the prominence of threat-oriented stressful events, adversity, and traumatic experiences in anxiety disorders, include a life history in the assessment. The cognitive case formulation should include appraisals of control, perceived vulnerability, and expected outcomes
associated with these experiences.
THE COGNITIVE VULNERABILITY MODEL
In their original account of the cognitive model of anxiety Beck et al. (1985, 2005)
defined vulnerability as “a person’s perception of himself as subject to internal or external dangers over which his control is lacking or is insufficient to afford him a sense of
safety. In clinical syndromes, the sense of vulnerability is magnified by certain dysfunctional cognitive processes” (Beck et al., 1985, pp. 67–68). In this formulation vulnerability to anxiety is conceptualized as a predisposition to misinterpret potentially
threatening or novel situations as dangerous and devoid of safety, leaving the individual
in a state of perceived helplessness. In the anxiety disorders only certain types of threat
will activate this underlying cognitive vulnerability. Once activated in a particular situation, the cognitive-affective program described in Chapter 2 (see Figure 2.1) maintains
the individual in a heightened state of anxiety.
Beck et al. (1985, 2005) focused on two main characteristics of cognitive vulnerability. The first is an enduring tendency to misinterpret certain types of threatening or novel
situations as dangerous. The second is a predisposition to perceive one’s self as incompetent, weak, or lacking the personal resources to deal with certain types of threatening or
stressful situations. In the current formulation of the cognitive model, the first feature of
cognitive vulnerability is captured by Hypothesis 12, enduring threat-related beliefs, and
112
COGNITIVE THEORY AND RESEARCH ON ANXIETY
the second falls under Hypothesis 11, elevated personal vulnerability. Both aspects of
vulnerability must be present for an individual to be cognitively predisposed to anxiety.
Furthermore, we would expect cognitive vulnerability to exhibit a high degree of selectivity within a diathesis– stress framework, so that it would only surface when the vulnerable person anticipates encountering specific types of potentially threatening situations.
Thus an enduring tendency to misinterpret certain types of potential threat and one’s
ability to manage this threat would remain dormant until activated by relevant trauma or
other forms of perceived stress. Once activated, the threat schemas would dominate the
information-processing system whenever a relevant threat-related cue is encountered.
Like other anxiety researchers, we believe a cognitive vulnerability for anxiety
develops through repeated experiences of neglect, abandonment, humiliation, and even
trauma that can occur during childhood and adolescence (see Barlow, 2002; Chorpita
& Barlow, 1998; Craske, 2003). Certain parenting practices such as overprotection,
restriction of independence and autonomy, preoccupation with potential danger, and
encouraging escape and avoidance in response to anxiety could all contribute to the
development of a cognitive vulnerability to anxiety. Although there is some empirical
evidence that supports this conjecture, much of it is based on retrospective assessment
of childhood experiences (McNally et al., 2001). Large community-based longitudinal
studies that begin in childhood are needed in order to determine the developmental
antecedents of cognitive vulnerability to anxiety.
The present account of cognitive vulnerability is consistent with the proposals of
other cognitive-behavioral researchers. M. W. Eysenck (1992), for example, proposed
a hypervigilance theory of anxiety in which individuals with high trait anxiety have an
attentional system that is oriented toward threat detection when they are in potentially
threatening situations or in a state of high anxiety. Craske (2003) suggested that both
negative affectivity and a threat-based style of emotional regulation (i.e., a response
to arousal and distress characterized by avoidance and danger-laden expectations) are
vulnerability factors for anxiety. Rachman (2004) noted that people may be primed to
detect threat cues and overlook or minimize safety information. Mathews and MacLeod
(2002) argued that attentional and interpretative biases for threat constitute a vulnerability to anxiety. And Wells (2000) proposed that enduring metacognitive beliefs (i.e.,
beliefs about one’s thoughts) about worry, judgments of cognitive confidence, and the
importance of monitoring one’s thought processes constitutes a vulnerability for emotional disorders.
Our focus on the cognitive basis of vulnerability to anxiety must be understood
within the context of other etiological factors such as biological and developmental
determinants, NA, trait anxiety, anxiety sensitivity, diminished personal control, and
the like. This broader view of vulnerability is represented in Figure 4.1.
Prepotent threat schemas and perceived personal vulnerability or weakness are more
specific cognitive constructs that directly reflect the slightly broader constructs of high
anxiety sensitivity, diminished personal control, and sensitivity to negative evaluation,
which in turn are related to broad traits of negative emotionality and high trait anxiety.
In this way vulnerability to anxiety disorders involves the interaction of multiple pathways emanating from constitutional, developmental, environmental, personality, and
information-processing domains. Based on this framework for vulnerability, we turn to
consider the empirical evidence for the two main components of the model: an enduring
sense of personal vulnerability and the presence of hypervalent threat schemas.
Vulnerability to Anxiety
113
Genetic
Predisposition
Biological
Determinants
Developmental
Learning
Experiences
BROADER
High Negative Affect or
Emotionality
High Anxiety
Sensitivity
Aversive
Life Events
FOCUS
Trait Anxiety
Sensitivity to
Negative
Evaluation
Hypervalent Threat
Schema Activation
Diminished
Personal Control
Heightened Sense of
Personal Weakness
and Reduced Safety
NARROWER
FOCUS
State of Anxiety
FIGURE 4.1. Cognitive vulnerability model of anxiety.
Hypothesis 11. Heightened Sense of Personal Vulnerability
Highly anxious individuals will exhibit lower self-confidence and greater perceived
helplessness in situations relevant to their selective threats compared to nonanxious
individuals.
Beck et al. (1985, 2005) considered diminished self- confidence and self-doubt an
important aspect of cognitive vulnerability to anxiety. For the person who suffers from
anxiety a self- confidence set is replaced by a vulnerability perspective. When in a vulnerability mode, individuals evaluate their own abilities and competence as inadequate
for dealing with a perceived threat. As a result, they become tentative or withdraw
from a situation in a self-protective manner. When a confident attitude is adopted, the
individual focuses on the positives in a situation, minimizes the dangers, and may even
assume a greater sense of personal control than when low self- confidence prevails (Beck
et al., 1985, 2005). Adopting a confident mode increases the probability of success in
114
COGNITIVE THEORY AND RESEARCH ON ANXIETY
a threatening situation, whereas dominance of the vulnerability mode is more likely to
lead to failure and reinforce individuals’ belief in their incompetence because it is associated with self-questioning, uncertainty, and a weak or tentative response in a challenging
situation. Bandura’s (1991) concept of low perceived self-efficacy, as well as uncontrollability and unpredictability, are distal vulnerability factors in anxiety proposed by other
researchers (e.g., Chorpita & Barlow, 1998; Schmidt & Woolaway-Bickel, 2006) that
are consistent with the cognitive vulnerability concept of diminished self- confidence for
selective types of perceived threats.
There are three assumptions about the nature of low self- esteem in anxiety. First,
the lack of self- confidence is highly specific to the anxious concerns of the individual.
Unlike depression, where we find a generalized negative view of the self, the lower selfworth in anxiety is only evident in situations relevant to the person’s anxious concerns.
For example, a client with a specific phobia about swallowing was discouraged and
disheartened about his ability to eat in the presence of others and yet felt very competent when performing in front of hundreds as an amateur comedian. Second, lack of
self- confidence will be a significant determinant of self-protective responses in anxious
situations such as escape and avoidance, and deficit performance in dealing with the
situation. And third, lack of self- confidence in responding to certain types of perceived
threat arises from early childhood and other learning experiences and so acts as a vulnerability factor for the later development of an anxiety disorder.
Empirical Evidence
The first criterion of vulnerability is sensitivity to the disorder in question. Anxious
individuals should exhibit less self- confidence in dealing with threatening situations
relevant to their anxiety state than nonanxious individuals. Like depression, presence of
anxiety disorders is characterized by a significant lowering of self-esteem (e.g., Ingham,
Kreitman, Miller, Sashidharan, & Surtees, 1986). In fact a connection between low
self-esteem and anxiety has figured prominently in psychological theories and research
on social anxiety, in particular. Various studies have shown that low self-esteem or
dysfunctional beliefs about the self are related to heightened social anxiety or shyness
(de Jong, 2002; Jones, Briggs, & Smith, 1986; Kocovski & Endler, 2000; Tanner et al.,
2006; Wilson & Rapee, 2006). However, there are a number of qualifications that must
be made about the nature of low self-esteem in social anxiety.
First, most of the research evidence indicates that the lack of self- confidence in
social phobia is specific to social situations involving the perception of evaluation from
others rather than a global low self- esteem. In fact social threat is often needed to prime
low self-worth in socially anxious samples (e.g., O’Banion & Arkowitz, 1977; Rapee
& Lim, 1992; Stopa & Clark, 1993). Second, it is not clear whether the lack of selfconfidence in social anxiety reflects an elevation in negative self- evaluation or a reduction in positive self-evaluation. Mansell and Clark (1999) found that a high social anxiety group recalled fewer positive trait adjectives but not more negative adjectives than a
low social anxiety group after giving a 2-minute videotaped speech (see de Jong, 2002;
Tanner et al., 2006, for similar findings). Thus the primary problem in social anxiety
may be a reduction in positive self- evaluation in social situations rather than an elevation in negative self-view. Third, it is still unclear which aspects of low self- esteem may
Vulnerability to Anxiety
115
be most important in social phobia. Wilson and Rapee (2006), for example, found that
it was certainty of self-concept that was reduced in social phobia, whereas Mansell and
Clark (1999) found that socially anxious individuals had reduced positivity recall for
public but not private self-referent trait adjectives. Finally, differences in self-esteem may
depend on whether automatic (i.e., implicit) or more effortful (i.e., explicit) processes are
assessed. Implicit Association Test (IAT) studies suggest that the problem of low selfesteem in anxiety may be reflected in more controlled, effortful processes rather than an
underlying, automatic evaluative bias (see de Jong, 2002; Tanner et al., 2006).
Although considerably less is known about the role of low self- esteem in other anxiety disorders, there is some preliminary research that is noteworthy. Ehntholt, Salkovskis, and Rimes (1999) found that both OCD and non-OCD anxious groups had significantly lower self-worth and generalized self- esteem than a nonclinical control group but
concluded that low generalized self- esteem may be a consequence of anxiety rather than
a predisposing factor. Wu, Clark, and Watson (2006) found that OCD patients were
distinguished by a very low self-image based on profile analysis of the SNAP-2 and low
self-esteem has been implicated in the development of PTSD symptoms (Piotrkowski &
Brannen, 2002). Doron and Kyrios (2005) proposed that a restricted self- concept may
constitute an underlying vulnerability for OCD. Thus there is increasing interest among
researchers on the role that low self-esteem and other selfhood concepts might play in
the pathogenesis of the anxiety disorders.
Summary
Although there is empirical evidence that low self- esteem characterizes the anxiety disorders, it is not clear whether this is a cause or a consequence of the disorder. Research
on self-esteem vulnerability in anxiety has lagged far behind the empirical literature
on low self- esteem in depression. Two types of studies are critical to progress beyond
mere speculation. First, longitudinal studies are needed to determine if low self-worth
is indeed a predisposing contributor to an anxiety disorder. These types of studies are
practically nonexistent in the anxiety literature. And second, experimental research is
needed to determine if variations in self- esteem have a corresponding causal effect on
anxious symptoms. Causal effects must be demonstrated if low self- confidence in dealing with threat is a true cognitive vulnerability for anxiety.
If low self-worth is a cognitive vulnerability for anxiety, the preliminary findings
suggest it is highly specific to threatening content perceived relevant to an individual’s
primary anxious concerns. In addition the lack of self- confidence is most likely evident
at the secondary phase of anxiety where effortful, controlled processes predominant
(see Figure 2.1). However, a conclusion on the empirical support for Hypothesis 11 must
wait until further research has been completed.
Clinician Guideline 4.4
When assessing self-worth issues in anxiety, the clinician should evaluate the client’s level of
self-confidence in dealing with situations that exemplify the individual’s primary anxious
concerns.
116
COGNITIVE THEORY AND RESEARCH ON ANXIETY
Hypothesis 12: Enduring Threat-Related Beliefs
Individuals vulnerable to anxiety can be distinguished from nonvulnerable persons by
preexisting maladaptive schemas (i.e., beliefs) about particular threats or dangers and
associated personal vulnerability that remain inactive until triggered by relevant life
experiences or stressors.
The cognitive model of anxiety (see Chapter 2) considers automatic activation of
the primal threat mode a central process in the experience of anxiety. Threat mode
activation sets in motion the symptoms that constitute a state of anxiety. Moreover, the
dysfunctional beliefs or schemas that comprise the primal threat model are personal
and quite idiosyncratic to each individual. They are primarily learned through various
positive or negative experiences of threat or danger that occurred to self or significant
others. As such they are enduring representations of threat, which in the anxiety disorders are often excessive, biased and maladaptive. These dysfunctional threat-related
schemas will result in exaggerated appraisals of the probability and severity of threat,
underestimate personal coping ability, and minimize the presence of safety (Beck et al.,
1985, 2005).
In the cognitive model threat-relevant schemas constitute the core cognitive vulnerability for anxiety. The threat schemas of the anxiously vulnerable person are not only
qualitatively different from those of the nonvulnerable person in terms of containing
misinformation and bias about particular threats, but they are also “prepotent” in that
a broader range of less intense stimuli will activate the schemas. For example, most
people feel some anxiety before giving a public address that reflects activation of beliefs
such “It is important that I do a good job” and “I expect the audience will be receptive.” However, the person vulnerable to social anxiety might feel intense anxiety when
asked a question in a work-related meeting because of activation of schemas like “I can’t
speak-up, people will notice that my voice is trembly,” “They’ll think there is something
wrong with me,” “They’ll assume I must have an anxiety problem—a mental illness.”
In comparison to the nonvulnerable person, the individual with social anxiety has more
extreme, exaggerated schemas that lead to an exaggerated appraisal of the danger. Also
notice that a much less threatening situation triggers the threat schemas of the socially
anxious person. In this way the schematic representations of threat in the vulnerable
person are prepotent or hypervalent, leading to more frequent and intense activation.
Unlike the nonvulnerable person, activation of certain threat schemas in the vulnerable
person will tend to capture much of the information-processing resources so that the
more constructive schemas become relatively inaccessible to the person.
Empirical Evidence
Is there any evidence that threat-relevant beliefs or schemas constitute an enduring cognitive predisposition for clinical anxiety states? We have already reviewed a considerable amount of empirical evidence that is consistent with a schema-based cognitive
vulnerability to anxiety. In the previous chapter numerous studies by MacLeod, Mogg,
Bradley, Mathews, and others found that nonclinical individuals with high trait anxiety
had an attentional processing bias for threat, especially under conditions of stress (see
reviews by Mathews & MacLeod, 1994, 2002, 2005; Mogg & Bradley, 1998). The con-
Vulnerability to Anxiety
117
clusion reached by Mathews and MacLeod (2002) is that high trait-anxious individuals
have a cognitive vulnerability to anxiety in the form of a lower threshold for switching
from an avoidance to a vigilant information-processing mode.
A second source of supportive evidence for a schematic vulnerability to anxiety
comes from the anxiety sensitivity and diminished control studies reviewed in this chapter. Although it would be inaccurate to describe the ASI as a beliefs measure, it does
assess appraisals that are based on a variety of preexisting beliefs about physical sensations and anxiety. For example, the ASI item “It scares me when I become short of
breath” would be based on a preexisting belief such as “I am putting myself at severe
risk of being unable to breath when I feel short of breath.” If high ASI scores predict elevated likelihood of subsequent anxiety, we can generalize from these findings to
the beliefs that underlie ASI appraisals as supportive evidence that these beliefs constitute vulnerability for anxiety. The same generalization can be made from the research
on diminished control and negative attributional style in anxiety. Certain preexisting
beliefs about lack of control over anticipated threats will underlie control perceptions,
making these beliefs an important element in the proposal that diminished sense of
personal control is a vulnerability factor in anxiety. To summarize, the notion of preexisting dysfunctional beliefs that predispose to anxiety is a common feature of many
cognitive theories of anxiety disorders (e.g., D. A. Clark, 2004; Ehlers & Clark, 2000;
Wells, 2000; Wells & Clark, 1997).
DYSFUNCTIONAL ANXIETY BELIEFS
In order to investigate the role of dysfunctional beliefs in the etiology of anxiety, specific
belief measures are needed that directly assess threat schema content. Unfortunately,
research in this area is not as well developed as the experimental studies on attentional
bias or the brief prospective diathesis– stress studies found in depression. Nevertheless,
we are beginning to see more research on the role of threat-relevant schemas and beliefs
in clinical anxiety.
In recent years there has been considerable research on the belief structure of OCD.
An international group of researchers called the Obsessive Compulsive Cognitions
Working Group (OCCWG) proposed six belief domains as constituting a cognitive vulnerability to OCD: inflated responsibility, overcontrol of thoughts, overimportance of
thoughts, overestimated threat, perfectionism, and intolerance of uncertainty (OCCWG,
1997). Definitions of these belief domains can be found in Table 11.3.
An 87-item self-report questionnaire, the Obsessive Beliefs Questionnaire (OBQ),
was developed to assess the six OCD belief domains. Later factor analysis indicated
it could be reduced to 44 items that assessed three belief dimensions: responsibility/
threat estimation, perfectionism/intolerance of uncertainty, and importance/control
of thoughts (OCCWG, 2005). Two large-scale multisite clinical studies based on the
87-item OBQ revealed that OCD patients scored significantly higher than other nonobsessional anxious and nonclinical comparison groups on OBQ Control of Thoughts,
Importance of Thoughts, and Responsibility subscales, in particular, and the six OBQ
belief scales correlated better with self-reported OCD measures than with the BAI or
BDI (OCCWG, 2001, 2003; see Steketee, Frost, & Cohen, 1998, for similar results).
However, the six OBQ subscales are highly intercorrelated and they have strong correlations with other non-OCD measures like the Penn State Worry Questionnaire. At pres-
118
COGNITIVE THEORY AND RESEARCH ON ANXIETY
ent the OBQ is probably the best measure of OCD beliefs, although certain weaknesses
are apparent in its construct validity.
It is also becoming increasing clear that only certain belief domains like responsibility, importance, and control of thoughts may be specific to OCD whereas other domains
like threat overestimation and perfectionism are common across the anxiety disorders.
Although there has been some inconsistency across studies, beliefs about the importance
of thoughts and need to control thoughts have tended to differentiate OCD patients
from other anxiety groups, with responsibility and overestimated threat sometimes
showing specificity but perfectionism and intolerance of uncertainty more often emerging as nonspecific across the anxiety disorders (e.g., Anholt et al., 2006; Clark, Purdon,
& Wang, 2003; Sica et al., 2004; Tolin, Worhunsky, & Maltby, 2006; see Emmelkamp
& Aardema, 1999, for contrary results). Moreover, some beliefs may be particularly
relevant for certain OCD subtypes such as importance/control of thoughts for pure
obsessions, or perfectionism/intolerance of uncertainty for OCD checking (Calarami et
al., 2006; Julien, O’Connor, Aardema, & Todorov, 2006). Also, cluster analytic studies with the OBQ suggest that not all patients with OCD will necessarily endorse these
OCD beliefs, leading some researchers to question whether dysfunctional beliefs plays a
role in all OCD cases (Calamari et al., 2006; Taylor et al., 2006).
Recently there has been an attempt to determine whether preexisting dysfunctional
beliefs might prospectively predict an escalation in OC symptoms. Eighty-five parents
who were expecting their first child were administered the OBQ-44 and other measures of anxious and obsessional symptoms prenatally and then 3 months postpartum
(Abramowitz, Khandker, Nelson, Deacon, & Rygwall, 2006). Most of the mothers
and fathers reported distressing intrusive thoughts about their newborns at the follow-up assessment, and regression analyses revealed that OBQ Total Scores predicted
an increase in postpartum OC symptoms as determined by the Yale–Brown Obsessive Compulsive Scale and the Obsessive– Compulsive Inventory— Revised. In a 6-week
prospective study involving 377 undergraduates, Coles and Horng (2006) found that
OBQ-44 Total Scores predicted an increase in OC symptoms as measured by the Obsessive Compulsive Inventory Total Score but the interaction between beliefs and negative
life events failed to reach significance. However, in a second study Coles and colleagues
failed to entirely replicate this finding (Coles, Pietrefesa, Schofield, & Cook, 2007), with
OBQ-44 showing only a trend toward significance and no interaction with negative life
events.
Researchers have examined the types of dysfunctional beliefs found in other anxiety disorders. Preexisting maladaptive beliefs about worry and its consequences are evident in chronic worry and GAD (Cartwright-Hatton & Wells, 1997; Dugas et al., 2005;
Dugas, Gagnon, Ladouceur, & Freeston, 1998; Wells & Cartwright-Hatton, 2004;
Ruscio & Borkovec, 2004; Wells & Papageorgiou, 1998a). Wenzel, Sharp, Brown,
Greenberg, and Beck (2006) found that beliefs relevant to panic such as the anticipation
of anxiety, concern about physical and emotional catastrophes, and self-deprecation
were more closely associated with anxiety and panic symptoms than with self-reported
depression. Individuals with social anxiety may endorse a number of early maladaptive
schemas as indicated by elevated scores on the Young Schema Questionnaire subscales
of Emotional Deprivation, Guilt/Failure, Social Undesirability/Defectiveness, Dependence, and the like (Pinto- Gouveia, Castilho, Galhardo, & Cunha, 2006). Overall there
is some indication that enduring maladaptive beliefs about threat and vulnerability
Vulnerability to Anxiety
119
characterize the anxiety disorders, but this research is still in its infancy and many
fundamental questions about the nature of schematic vulnerability in anxiety have not
been addressed.
INDUCED THREAT INTERPRETATION BIAS
It is now well established that a tendency to endorse threatening interpretations of ambiguous information is an important feature of the selective processing bias for threat that
characterizes anxiety (Mathews, 2006). However, demonstrating that threat-processing
bias, and its underlying schematic threat activation by extension, has causal influence
is more difficult because most of the research has been correlational or involved crosssectional research designs. Mathews and MacLeod (2002) note that evidence of differential bias in anxious and nonanxious groups, reduction of threat bias with treatment, or differential activation of bias in high and low trait anxious individuals after
a stressful event can not rule out a noncausal explanation such as the influence of a
third unidentified variable. Thus research showing that experimental manipulation of
interpretative bias through deliberate training conditions has a considerable impact on
emotion is strong empirical evidence for causality in evaluative processing of threat.
Furthermore, this research is important for cognitive vulnerability because it provides
evidence for a basic precondition of vulnerability: that biased information processing
has a causal effect on emotion.
The basic aim of induction procedures is to train volunteers to engage in selective
processing of new anxiety-relevant information and assess changes in subsequent anxiety. Two effects are necessary to demonstrate. First, that training in differential processing bias has been successful and generalizes to the processing of new information. And
second, an increase or decrease in threat-processing bias results in changes in level of
anxiety. A third question often addressed is whether there are individual differences
in susceptibility to threat-bias training that might suggest heightened vulnerability to
anxiety.
MacLeod and colleagues conducted a series of experiments on induced attentional
bias for threat in student volunteers. In the typical experiment individuals were randomly
assigned to an attentional threat training condition or the avoidance of threat in favor of
emotionally neutral cues (Mathews & MacLeod, 2002). In a series of unpublished pilot
experiments (see discussion in Mathews & MacLeod, 2002), MacLeod and colleagues
adapted the dot probe detection paradigm so that participants were randomly assigned
to 576 training trials in which the dot always appeared in the location of threatening or
neutral words. Analysis of 128 test trials revealed a significant training effect in which
participants trained to detect threat words were significantly faster at probe detection
after a threat word and slower to detect probes after a neutral word. This training effect
was replicated in another pilot experiment using happy and angry faces.
In their first major published study, MacLeod et al. (2002) reported on two studies
involving experimental manipulation of attentional bias. In the first experiment 64 nonvulnerable students (trait anxiety scores in the middle range) were randomly assigned
to an “attend negative” training condition or an “attend neutral” condition. Training
involved 576 trials in which 50% of the word pair presentations were at a short exposure
interval (i.e., 20 milliseconds) and the other 50% were at a longer exposure duration
(i.e., 480 milliseconds). Ninety-six test trials were distributed throughout the training
120
COGNITIVE THEORY AND RESEARCH ON ANXIETY
trials. Thus half of the participants were trained to attend to the negative information
and the other half were trained to attend away from negative stimuli (attend to neutral
words). After the dot probe training all participants completed a stressful anagram
task. Analysis revealed that students in the negative training condition exhibited faster
dot probe detection to negative words in the test trails, whereas participants trained to
attend away from negative words exhibited a speeding effect to dot probes following
the neutral words. However, this training effect was only evident at the longer exposure
trials, indicating that differential bias was not preconscious. Furthermore, attentional
training had no immediate effect on mood, although after the anagram stress students
trained to attend away from negative information showed significantly lower elevations
in negative mood. The authors concluded that attentional threat avoidance training
may reduce vulnerability for negative emotional response to stress.
In a second replication study all training trials were conducted at a longer exposure
interval and emotional reactivity to stress was assessed before and after attentional
training (MacLeod et al., 2002). Analysis revealed that a differential training effect was
again achieved and that attentional training away from negative stimuli resulted in no
negative emotional response to the anagram stressor, whereas the group that had negative attentional training showed a pronounced negative emotional response to the stressor. These differential effects were due to training because at baseline the groups did
not differ in showing elevations in negative mood to a preinduction baseline anagram
task. The authors concluded that attentional training modified the degree of emotional
response to a subsequent stressor. Thus the training had its greatest impact not on mood
directly but rather on affecting emotional vulnerability to stress.
Of greatest relevance to Hypothesis 12 are a series of published studies on interpretative bias training. Grey and Mathews (2000) first investigated whether interpretative bias for threat could be trained in volunteers with normal trait anxiety scores.
Individuals were randomly assigned to a threatening or a nonthreatening homograph
training condition in which volunteers were trained to complete a word fragment with a
threatening or nonthreatening homograph. In the first experiment, Grey and Mathews
(2000) found that threat training resulted in faster response for generating threat solutions on 20 critical test items, and the biasing effect of threat training was found to
generalize to a lexical decision task in two further experiments. In a final study that
included an untrained control group, individuals exposed to homograph threat training
showed faster lexical decision for threat than the baseline group. These studies, then,
demonstrated that an interpretative threat bias for ambiguous stimuli can be trained in
nonvulnerable individuals.
Mathews and Mackintosh (2000) conducted five experiments in which interpretative bias training involved making a negative (threatening) or positive (nonthreatening) interpretation to a short description of an ambiguous social situation. Sixty-four
descriptions were presented with each one followed by a word fragment that matched
a threatening or nonthreatening interpretation. In the first experiment, volunteers randomly assigned to interpretative threat training were faster at completing negative probe
word fragments and gave higher recognition ratings to threatening interpretations of the
ambiguous descriptions. Furthermore, there was a direct effect on mood, with the threat
group reporting an increase in anxiety after training, although this mood effect was
not replicated in the second experiment. In the fourth experiment interpretative threat
training did result in an increase in state anxiety but its effects were shown to dissipate
Vulnerability to Anxiety
121
rather rapidly. The final experiment demonstrated that induced bias for threat will lead
to an increase in anxiety only when it is activated by generating personally threatening
meanings. The authors conclude that their results provide direct experimental evidence
that activation of threat interpretation bias plays a causal role in anxiety.
In a more recent study Wilson et al. (2006) used the homograph interpretative bias
induction of Grey and Mathews (2000) and randomly assigned 48 nonanxious students
to a threat or nonthreat training condition. Analysis revealed the expected differential
interpretation bias with training but no direct effect on depressed or anxious mood.
However, interpretation bias did have a significant impact on emotional reactivity to
four stressful video clips with the threat-trained group showing an elevation in state
anxiety in response to the stressor. The authors concluded that threat interpretation bias
can make “a causal contribution to anxiety reactivity” (Wilson et al., 2006, p. 109).
Yiend, Mackintosh, and Mathews (2005) used the text-based ambiguous social
scenarios from Mathews and Mackintosh (2000) to demonstrate that the induction of
a threat interpretation bias can endure over at least 24 hours but, like previous studies,
there was no significant direct effect on state anxiety. In another study Mackintosh,
Mathews, Yiend, Ridgeway, and Cook (2006) again found that induced interpretation
bias endured over a 24-hour time period and survived changes in environmental context
between training and testing. This enduring effect of induction training was replicated
in a second experiment using text-based scenarios involving potential physical threat.
Furthermore, individuals with the negative interpretation training showed the largest
increases in state anxiety after viewing stressful accident video clips a day after training.
However, a replication study of Mathews and Mackintosh (2000) failed to find that the
effects of interpretative bias training generalized to indices of interpretative processing that differed from the training task, although they did find that negatively trained
individuals had significant increases in state anxiety (Salemink et al., 2007a). A second
experiment, however, produced negative results, with positive and negative interpretative bias training having no significant effect on state anxiety or emotional reactivity
to stress (Salemink et al., 2007b). Together these results indicate that interpretative
training effects can endure over time and across differing environmental and possibly
stimulus contexts, and that changes in emotional reactivity due to training may also
have some measure of durability.
In a special issue of the Journal of Abnormal Psychology a series of studies based
on cognitive bias training demonstrated that significant therapeutic benefits could be
achieved from directly training anxious individuals to generate benign or positive interpretations to emotionally ambiguous material, or to selectively attend to nonthreatening
stimuli; the procedures were labeled cognitive bias modification (for a discussion see
MacLeod, Koster & Fox, 2009). Four studies are of particular relevance in demonstrating the causal status of threat bias. In the first study nonclinical students who were
trained over several days to selectively avoid emotionally negative or threatening words
using a home-based dot probe program had significantly lower trait anxiety scores and
weakened stress reactivity to a naturalistic stressor encountered 48 hours after training
than a no-train control group (MacLeod & Bridle, 2009).
In a second study high worriers trained to access benign meanings to threat-related
homographs and emotionally ambiguous scenarios had significantly fewer negative
thought intrusions and less anxiety during a focused breathing task than the no-training
control group (Hirsch, Hayes, & Mathews, 2009). In two final studies involving atten-
122
COGNITIVE THEORY AND RESEARCH ON ANXIETY
tional training using a dot probe task, individuals with GAD trained to selectively attend
to neutral words had a significant decrease in attentional threat bias and anxiety symptoms (Amir, Beard, Burns, & Bomyea, 2009), and in a second similar study socially
anxious participants trained to disengage from negative social cues also reported significantly greater reductions in social anxiety and trait anxiety than the no-training control
group (Schmidt, Richey, Buckner, & Timpano, 2009). Together these studies indicate
that cognitive bias training may be effective in reducing anxiety, which provides further
support for a causal basis to threat bias in anxiety.
Summary
There is relatively little research on cognitive vulnerability to anxiety that has employed
self-report questionnaires of dysfunctional beliefs about threat, except for some studies reporting inconsistent findings on enduring beliefs in OCD. However, more recent
experimental studies employing different training protocols have demonstrated that a
threat interpretation bias can be created in nonanxious individual that may be similar
to the selective processing bias for threat that characterizes anxiety. Evidence of some
durability over time and transfer of induced processing style to novel stimuli and changes
in environmental context suggests that these training effects may be quite robust. However, the causal effects of induced threat interpretation bias on anxiety are not simple.
It is apparent that training effects on anxiety are most likely when the induced bias is
activated when individuals are required to generate personally threatening meanings
(Mathews & Mackintosh, 2000) or, possibly when interpretation bias activates personally threatening imagery (Hirsch, Clark, & Mathews, 2006). Moreover, the moodcongruency effects of induced interpretation bias are most notable with exposure to a
stressor. Thus the evidence to date indicates that threat interpretation bias plays a causal
role in modifying vulnerability to emotional reactivity. However, this research is still in
its infancy and many fundamental questions remain unanswered.
Training in positive interpretation bias may prove to be an effective treatment for
clinical anxiety states. Studies on cognitive bias modification have demonstrated significant reductions in anxious symptoms. Mathews et al. (2007) found that training in
positive interpretation bias reduced trait anxiety scores. Furthermore, use of imagery
during interpretation training might improve training effects as indicated by reductions
in state anxiety and increases in positive affect (Holmes, Mathews, Dalgleish, & Mackintosh, 2006; see also Holmes, Arntz, & Smucker, 2007). The current findings, then,
are most promising and are our strongest experimental evidence to date that schematic
threat activation in the form of interpretative threat bias plays a significant contributory role in anxious reactivity to stress. Moreover, there may be significant therapeutic
benefits in reversing the preexisting cognitive bias by training vulnerable individuals to
make positive interpretations of ambiguous threat stimuli.
Clinician Guideline 4.5
Deliberate and sustained training in generating positive, nonthreatening interpretations of
personally meaningful situations relevant to the client’s primary anxious concerns can counter the hypervalent schematic threat activation that characterizes vulnerability to anxiety.
Vulnerability to Anxiety
123
SUMMARY AND CONCLUSION
In this chapter we discussed a number of constructs that have been proposed in the etiology of anxiety disorders. Although various genetic, biological, developmental, and environmental factors have been implicated in the onset of anxiety, it is our contention that
individuals can also possess cognitive vulnerability for anxiety. As depicted in Figure
4.1, the cognitive model recognizes that genetic predisposition, biological determinants,
childhood experiences, and aversive life events all play a significant role in the etiology
of an anxiety disorder. At the same time, however, general cognitive-personality factors
interact with more specific enduring cognitive structures as contributory pathways to
the expression of anxiety.
At the more general level the cognitive model recognizes that certain personality
characteristics such as high negative emotionality or elevated trait anxiety are nonspecific vulnerability factors in anxiety. There is now considerable empirical evidence
that nonclinical high trait-anxious individuals exhibit a propensity for a threat-related
information-processing bias that is similar to that seen in the anxiety disorders, especially when induced by training or activated by a stressor (e.g., see review by MacLeod
et al., 2004). High NA has been implicated in the etiology of both anxiety and depression. However, it is at the more specific level that we see contributory factors that have
even more relevance for anxiety. An extensive literature now exists on the etiological
role of anxiety sensitivity and while perceived uncontrollability is clearly involved in the
pathogenesis of anxiety, it is doubtful its influence is limited to the anxiety disorders.
The remainder of the chapter discussed evidence for the final two hypotheses of
the cognitive model. There is emerging evidence that enduring beliefs or schemas about
threat and personal vulnerability are predisposing factors to anxiety. Although research
on a cognitive vulnerability model of anxiety is still in its infancy, considerable progress
has been made in the last few years in demonstrating the causal status of an informationprocessing bias for threat in anxiety. We are only beginning to see how this cognitive
vulnerability research might lead to better treatments for the anxiety disorders.
PART II
COGNITIVE THERAPY OF ANXIETY
Assessment and Intervention Strategies
The reformulated generic cognitive model of anxiety presented in Part I pro-
vides a framework for assessment and case formulation as well as for cognitive
and behavioral approaches to intervention that are common across the anxiety
disorders. In this sense cognitive therapy is transdiagnostic, targeting maladaptive cognitive structures and processes that are common across the various subtypes of anxiety. The chapters in this part of the book provide detailed, stepby-step instructions for basic cognitive assessment and treatment approaches
that are relevant to all forms of anxious symptom presentation. Chapter 5 discusses standardized measures for assessing general anxiety as well as a framework and case illustration for developing a cognitive case formulation for anxiety. Chapter 6 explains how to implement cognitive intervention strategies like
education, self-monitoring, cognitive restructuring, and generating alternatives
to modify the exaggerated threat and vulnerability appraisals and beliefs in
anxiety disorders. Chapter 7 focuses on the critical role played by behavioral
interventions such as exposure, response prevention, and directed behavioral
change in cognitive therapy for anxiety disorders. Together these chapters provide basic instruction in how to implement core cognitive and behavioral intervention strategies that provide the scaffolding for the disorder-specific cognitive therapy discussed in Part III.
125
Chapter 5
Cognitive Assessment
and Case Formulation
Our Age of Anxiety is, in great part, the result of trying to do today’s jobs
with yesterday’s tools.
—M ARSHALL MCLUHAN (Canadian academic and author, 1911–1980)
Sharon is a 52-year-old single woman who worked as an information technology consultant for a large advertising firm. She had been employed with this
firm for 10 years, and her job involved daily contact with a large number of
employees who requested her assistance whenever they experienced problems
with their computers. Thus her job required many daily one-to-one interactions with individuals at their workstations dealing with their computer and
network problems as well as meetings with senior managers whenever there
were questions about information technology.
Sharon decided to finally seek treatment for what she described as a “lifelong struggle with anxiety.” She indicated that her main problem was heightened anxiety whenever she engaged in social interaction with work colleagues.
She reported only mild anxiety outside the work setting and so never before
considered treatment until 6 months ago when she experienced a significant
increase in her work setting anxiety level. She declined pharmacotherapy from
her family physician and instead agreed to see a psychologist for psychotherapy.
Before offering Sharon a course of cognitive therapy, there were a number of
questions about her anxiety that needed to be addressed. What was the nature
of her anxiety disorder and what were her primary anxiety symptoms? What
external or internal cues triggered her anxiety? What were her automatic anxious thoughts and exaggerated appraisals of threat and personal vulnerability?
Was she highly intolerant of anxiety and hypervigilant for certain symptoms
of anxiety? How did she try to cope with her heightened anxiety? Were worry
and avoidance prominent responses to anxiety? How did she interpret her failure to control anxiety? These are a few of the questions that were addressed
during Sharon’s assessment sessions that led to an individualized cognitive case
formulation that is presented at the end of this chapter.
127
128
ASSESSMENT AND INTERVENTION STRATEGIES
Assessment and case formulation stand as a bridge between cognitive theory and
treatment. Since its earliest inception cognitive therapy has emphasized the importance
of theory-guided assessment as the foundation for effective psychotherapy. In the first
published cognitive therapy manual, Beck, Rush, Shaw, and Emery (1979) emphasized
that diagnostic formulation, establishing treatment goals, educating the client into the
cognitive model, and selecting target symptoms were critical elements in treatment for
depression.
The tools of assessment and case formulation that are now available to the cognitive therapist are much more precise than those available in the early years of cognitive
therapy. For example J. S. Beck (1995, 2005) developed a more detailed and refined
cognitive case conceptualization scheme that can be applied to the anxiety disorders.
She argues for the importance of conceptualization as a guide for focusing therapy on
the critical problems and processes that underlie a psychological disturbance. Often
treatment failure in difficult cases can be traced to a misguided or incomplete case
conceptualization (J. S. Beck, 2005). Persons and colleagues (Persons, 1989; Persons
& Davidson, 2001) provided one of the most comprehensive models for case formulation, emphasizing its individualized, theory-driven, and hypothesis-generating nature.
Cognitive-behavioral treatment protocols for specific anxiety disorders like panic (S.
Taylor, 2000), social phobia (Elting & Hope, 1995), GAD (Turk, Heimberg, & Mennin,
2004; Wells, 1997) and OCD (D. A. Clark, 2004) again emphasize the important role
played by cognitive assessment and case formulation.
In this chapter we present a case formulation scheme for anxiety based on the cognitive model (see Figure 2.1). A general framework for cognitive case conceptualization
is described that can be applied to all anxiety disorders. Precise applications of this case
conceptualization scheme will be considered within the disorder-specific chapters. The
first section of the chapter reviews diagnostic and general anxiety symptom measures
that are an important assessment tool in cognitive therapy of anxiety. This will be followed by a discussion of the assessment of immediate fear activation (Phase I) and its
sequelae. A third section focuses on assessment of secondary, elaborative processes that
lead to a reappraisal of threat and personal vulnerability. The chapter concludes with a
case illustration of cognitive formulation of anxiety and a consideration of difficulties
that can arise at this stage of treatment.
DIAGNOSTIC AND SYMPTOM ASSESSMENT
The first two or three contact sessions should focus on assessment that leads to a preliminary case formulation. Figure 5.1 illustrates a three-pronged approach to assessment
that will be present during the initial phase of cognitive therapy for anxiety.
Diagnostic Interviews
The diagnostic interview has always played an important role in cognitive therapy. Beck
et al. (1979) argued that a complete diagnostic evaluation is essential for establishing
symptom targets and treatment planning. Although clinicians are divided on the importance of differential diagnosis in psychotherapy, there is no debate that critical clinical
Cognitive Assessment and Case Formulation
129
Diagnostic
Information
Symptom
frequency
and severity
Personal
(idiographic)
data
FIGURE 5.1. Three aspects of assessment for anxiety.
information is obtained in the course of conducting a diagnostic interview. A diagnostic
interview is important to case conceptualization and treatment planning because:
•฀ It provides detailed information on the presenting symptom typology, frequency,
and severity.
•฀ Key cognitive processes in the anxiety disorders are often assessed.
•฀ Situational triggers and coping strategies, especially avoidance responses, are
evaluated.
•฀ Level of distress and impact on daily functioning is determined.
•฀ Precipitating factors, symptom development, and course are delineated.
•฀ Competing symptoms and other psychological processes that might complicate
treatment are identified.
Two key questions must be settled before conducting a diagnostic assessment in
cognitive therapy. Is it really necessary to spend the extra time doing a structured or
semistructured clinical interview, or would a traditional unstructured interview suffice?
Which is the best structured diagnostic for the anxiety disorders? Experts agree that
structured or semistructured interviews must be used to establish diagnostic status in
clinical research (Antony & Rowa, 2005). This is because structured interviews are
significantly more accurate in determining a valid diagnosis than unstructured clinical
interviews (Miller, Dasher, Collins, Griffiths, & Brown, 2001), and they have greater
interrater reliability (Miller, 2001). Miller (2002) determined that the diagnostic imprecision of traditional unstructured clinical interviews was in large part due to incomplete
data collection. Because semistructured interviews force the clinician to assess all key
diagnostic symptoms, this error in data collection is overcome.
Despite the diagnostic superiority of semistructured interviews, they are rarely used
in clinical practice (Antony & Rowa, 2005). This is because semistructured interviews
130
ASSESSMENT AND INTERVENTION STRATEGIES
can take upwards of 2 hours to administer, they require some degree of training, and
the published booklets can be quite costly. Nevertheless, we believe that the wealth of
information obtained from an interview like the Anxiety Disorders Interview Schedule
for DSM-IV (ADIS-IV) or the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-IV) justifies the investment in clinical resources (see Miller, 2002, for cost–
benefit analysis).
Although a fairly wide selection of interviews is available to the clinician, the ADISIV (Brown, Di Nardo, & Barlow, 1994) and SCID-IV (First, Spitzer, Gibbon, & Williams, 1997) have become the most widely used interviews in North America. Both
are clinician-administered, semistructured interviews designed to make a differential
diagnosis based on DSM-IV-TR criteria (APA, 2000). The SCID for Axis I has a published clinician version (SCID-CV) that covers the DSM-IV-TR diagnoses most commonly seen in clinical practice, whereas the unpublished research version (SCID-RV)
is much longer and includes numerous diagnostic subtypes and course specifiers (First
et al., 1997). Summerfeldt and Antony (2002) concluded that the SCID is superior in
its breadth of diagnostic coverage and there is evidence of good interrater reliability for
many of the most common diagnostic disorders (Williams et al., 1992; Riskind, Beck,
Berchick, Brown, & Steer, 1987). However the SCID-CV provides only a brief symptom
screener for certain anxiety disorders like specific phobia, GAD, social phobia, and agoraphobia without a history of panic disorder while failing to assess past history of other
disorders. In order to obtain an accurate diagnosis of specific anxiety disorders, the
SCID-CV must be supplemented with additional symptom questions from the SCIDRV. The addition of dimensional severity ratings on situational triggers is also recommended in order to provide important clinical data on the specific anxiety disorders
(Summerfeldt & Antony, 2002).
The best diagnostic interview for the anxiety disorders is the ADIS-IV. Although
the ADIS-IV has current and lifetime versions available for adults, the current version
will be of most relevance in clinical practice. It includes sections on each of the anxiety
disorders as well as highly comorbid conditions (e.g., mood disorders, hypochondriasis,
alcohol/drug abuse or dependence). In each of the anxiety disorder sections, severity
and distress ratings are obtained on specific symptoms, and the Hamilton Rating Scale
of Anxiety (HRSA; Hamilton, 1959) and Hamilton Rating Scale of Depression (HRSD;
Hamilton, 1960) are included so that the scales can be administered during the interview. Although the ADIS-IV covers all the key diagnostic criteria for the anxiety disorders, it goes well beyond DSM-IV-TR by providing information on psychopathological
phenomena that are targeted in interventions for anxiety (e.g., partial symptom expression, avoidance, situational triggers, and apprehension).
The ADIS-IV has high interrater reliability for the DSM-IV-TR anxiety and mood
disorders (see review by Summerfeldt & Antony, 2002). Brown and Barlow (2002)
reported that the ADIS-IV current or lifetimes versions had good to excellent interrater agreement for principal diagnoses based on a clinical sample of 362 outpatients
(see also Brown, Di Nardo, Lehman, & Campbell, 2001). Kappas for two independent interviews conducted within a 2-week interval ranged from .67 for GAD to .86
for specific phobia. The most common source of disagreement among the interviewers
involved whether a case met threshold criteria for a particular anxiety disorder as well
as information variance across interviews (i.e., patients giving different information to
the interviewers). Summerfeldt and Antony (2002) noted that although the ADIS-IV
Cognitive Assessment and Case Formulation
131
provides more detailed information and dimensional ratings of anxious symptoms, it is
more time-consuming and assesses a narrower range of disorders. The ADIS-IV can be
purchased from Oxford University Press/Graywind Publications.
Clinician Guideline 5.1
Administer the ADIS-IV current version prior to implementing a course of cognitive therapy
for anxiety. The ADIS-IV provides a precise diagnosis and crucial symptom data for the five
anxiety disorders discussed in this volume.
Symptom Measures
A number of standardized self-report questionnaires and clinician rating scales are
available to assess the frequency and severity of anxious symptoms. Here we focus on
broadly based, general measures of anxiety with disorder-specific measures covered in
later chapters. Standardized measures of general anxiety symptoms are useful because
they provide:
•฀ A broad overview or screening of various anxious symptoms.
•฀ A measure of symptom severity that is important for evaluating treatment effectiveness.
•฀ Access to normative data so that the relative severity of an anxiety state can be
determined.
•฀ Opportunity for repeated administration over the course of treatment so that
progress can be charted and symptom clusters identified that have been unresponsive to treatment.
Over the years a variety of general anxiety measures have been developed. The following section presents a few measures that we believe are most relevant for cognitive
therapy of anxiety. A more comprehensive review of anxiety measures is provided in an
edited book by Antony, Orsillo, and Roemer (2001).
Beck Anxiety Inventory
The Beck Anxiety Inventory (BAI; Beck & Steer, 1990) is a 21-item questionnaire that
assesses the severity of anxious symptoms on a 0 (“not at all”) to 3 (“severely, I could
barely stand it”) scale. According to the manual (Beck & Steer, 1990), the normal range
for the BAI Total Score is 0–9, mild anxiety is 10–18, moderate severity is 19–29, and
severe anxiety ranges from 30 to 63. Psychometric studies indicate that the BAI has
high internal consistency (alpha = .92) and a 1-week test– retest reliability of .75 (Beck,
Epstein, Brown, & Steer, 1988; Steer, Ranieri, Beck, & Clark, 1993). The BAI Total
Score correlates moderately with other anxious symptom measures like the Hamilton
Rating Scale of Anxiety— Revised, State–Trait Anxiety Inventory, and weekly diary
ratings of anxiety, and patients with anxiety disorders score significantly higher than
those with other psychiatric diagnoses (Beck et al., 1988; Creamer, Foran, & Bell, 1995;
Fydrich, Dowdall, & Chambless, 1992; Steer et al., 1993). As reported in the manual
132
ASSESSMENT AND INTERVENTION STRATEGIES
(Beck & Steer, 1990), the BAI Total Score means and standard deviations for various
diagnostic groups are as follows: panic disorder with agoraphobia (M = 27.27, SD =
13.11), social phobia (M = 17.77, SD = 11.64), OCD (M = 21.69, SD = 12.42), GAD (M
= 18.83, SD = 9.08), and primary depressive disorder (M = 17.80, SD = 12.20).1 Factor
analyses indicate that the questionnaire is multidimensional with either a two or a four
factor structure (e.g., Creamer et al., 1995; Hewitt & Norton, 1993; Steer et al., 1993).
However, only one-quarter of the items assess the subjective or more cognitive aspects
of anxiety (e.g., fear of the worst, unable to relax, terrified, nervous, scared) with the
remainder assessing the physiological hyperarousal symptoms of anxiety. Thus the BAI
is a good measure of the physical aspects of anxiety (especially panic disorder) and it is
sensitive to treatment effects, although like most anxiety measures it correlates highly
with self-report depression instruments (e.g., D. A. Clark, Steer, & Beck, 1994). The
BAI is available from Pearson Assessment at pearsonassess.com.
Hamilton Rating Scale of Anxiety
The Hamilton Rating Scale of Anxiety (HRSA; Guy, 1976; Hamilton, 1959) is a 14-item
clinician rating scale that assesses the severity of predominantly biological and behavioral symptoms of anxiety. Each symptom is rated on a severity scale from 0 (“not present”) to 4 (“very severe/incapacitating”) with symptomatic descriptions for each item.
A cut-off score of 14 on the HRSA Total Scale differentiates individuals with an anxiety disorder from those with no current diagnosis (Kobak, Reynolds, & Greist, 1993).
The HRSA Total Score has good internal consistency, interrater reliability, and 1-week
test–retest reliability, and it has strong convergent and discriminant validity as well as
sensitivity to treatment (Maier, Buller, Philipp, & Heuser, 1988; Moras, Di Nardo, &
Barlow, 1992; see review by Roemer, 2001). However, the majority of individuals with
major depression score above the cut-off score so the instrument does not accurately
discriminate anxiety from depression (Kobak et al., 1993). Given that some training is
required for the HRSA, the measure could be reserved for cases where a self-assessment
of anxiety might be highly inaccurate (i.e., individuals who minimize or exaggerate
their anxiety). A copy of the HRSA can be found in Appendix B of Antony et al. (2001)
or in the appendix of the ADIS-IV.
Depression Anxiety Stress Scale
The Depression Anxiety Stress Scale (DASS; Lovibond & Lovibond, 1995a, 1995b) is
a 42-item questionnaire with 14 items each assessing the severity of anxiety, depression, and stress. The anxiety subscale assesses autonomic arousal, skeletal musculature, situational, and subjective aspects of anxiety. For the DASS Anxiety Scale, 0–7
represents the normal range, 8–9 is mild anxiety, 10–14 is moderate, 15–19 is severe,
and 20+ is extremely severe (see Lovibond & Lovibond, 1995b). The subscale has
good internal consistency, temporal reliability, and convergent validity (Antony, Bieling, Cox, Enns, & Swinson, 1998a; Brown, Chorpita, Korotitsch, & Barlow, 1997;
1 The
BAI Total Score mean for the primary depressive disorder group (major depression, dysthymia, and
adjustment disorder with depressed mood) was derived from an intake data set (N = 293) from the Center
for Cognitive Therapy, University of Pennsylvania Medical School, that was available to the first author.
Cognitive Assessment and Case Formulation
133
Lovibond & Lovibond, 1995a). For example, DASS Anxiety correlates .81 with the
BAI and DASS Depression correlates .74 with the BDI in student samples (Lovibond
& Lovibond, 1995b). In addition, individuals with panic disorder score significantly
higher on DASS Anxiety than patients with major depression but those with OCD,
social phobia, GAD, and simple phobia do not score higher than the major depression
group (Antony, Bieling, et al., 1998; Brown et al., 1997). A shorter 21-item version
of the DASS was developed by Antony, Bieling, and colleagues (1998) and has psychometric characteristics comparable to the original 42-item DASS. Although DASS
Anxiety and Depression are moderately correlated (r’s ~ .45) in clinical samples and
DASS Anxiety has a predominant emphasis on autonomic arousal and fear (Antony,
Bieling, et al., 1998; Brown et al., 1997), it is a promising measure. The DASS-42 is
available in Appendix B of Antony et al. (2001) or it can be downloaded directly from
www.psy.unsw.edu.au/dass. The manual and scoring template can be ordered from
the same website.
State–Trait Anxiety Inventory
The State–Trait Anxiety Inventory (STAI—Form Y; Spielberger, Gorsuch, Lushene,
Vagg, & Jacobs, 1983) consists of two 20-item scales with one scale assessing state
anxiety (“how you feel right now, that is, at this moment”) and the other measuring trait anxiety (“how you generally feel”). With its emphasis on current state, the
STAI State scale has greater clinical relevance for tracking the effectiveness of cognitive therapy. Although the STAI has good reliability and convergent validity with other
anxiety measures, its ability to distinguish anxiety from depression has been questioned
(Roemer, 2001). For this reason we believe there are other anxiety symptom measures
that provide a clearer assessment for the cognitive therapist. The STAI—Form Y can be
purchased from Consulting Psychologists Press, Inc.
Cognitions Checklist
The Cognitions Checklist (CCL; Beck, Brown, Steer, Eidelson, & Riskind, 1987) comprises a 12-item anxiety subscale (CCL-A) and a 14-item depression subscale (CCL-D)
that assesses the frequency of negative self-referent anxious and depressive thoughts
along a 5-point scale ranging from 0 (“never”) to 4 (“always”). The content of CCL-A
revolves around themes of uncertainty and an orientation toward the future (Beck et
al., 1987), with the majority of items (71%) focused on anxious thinking about physical
or health-related concerns. Both subscales have good internal consistency, and factor
analyses reveal the expected loadings of CCL items on separate anxiety and depression
dimensions, especially in clinical samples (Beck et al., 1987; Steer, Beck, Clark, & Beck,
1994). Although CCL-A and CCL-D are moderately correlated, each subscale is more
highly correlated with its congruent than incongruent symptom state (Beck et al., 1987;
D. A. Clark et al., 1996; Steer et al., 1994). In clinical practice the CCL-A provides an
estimate of the frequency of anxious thoughts, especially the physical or health concerns
of most relevance to panic disorder. Individuals with anxiety disorders typically score in
the midteens or higher on CCLA (Steer et al., 1994). A copy of the CCL can be obtained
from the Center for Cognitive Therapy, Department of Psychiatry, University of Pennsylvania Medical School, Philadelphia, PA.
134
ASSESSMENT AND INTERVENTION STRATEGIES
Penn State Worry Questionnaire
The Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec,
1990) is a 16-item trait measure that assesses the propensity to worry as well as the
intensity of worry experiences without reference to specific worry topics (Molina &
Borkovec, 1994). The items are rated on a 5-point Likert scale from 1 (“not at all typical”) to 5 (“very typical”), with items 1, 3, 8, 10, and 11 reverse scored. Although there
is some debate over the factorial structure of the PSWQ (Brown, 2003; Fresco, Heimberg, Mennin, & Turk, 2002), only the Total Score is normally interpreted. The PSWQ
has high internal consistency, test–retest reliability, and correlates with other self-report
worry measures but it does have lower convergence with measures of general anxiety
(Brown, Antony, & Barlow, 1992; Davey, 1993; Meyer et al., 1990; Molina & Borkovec, 1994). Group comparisons indicate that individuals with GAD score highest on
the PSWQ, followed by other anxiety disorder groups and major depression who have
similar elevated scores that are significantly higher than nonclinical controls (Brown et
al., 1992; Chelminski & Zimmerman, 2003). A PSWQ cutoff score of 45 can be used to
identify pathological worry or GAD in a treatment-seeking population (Behar, Alcaine,
Zuellig, & Borkovec, 2003), although a higher cutoff score (62 or even 65) is needed
to differentiate GAD from other anxiety disorders and possibly even depression (e.g.,
Fresco, Mennin, Heimberg, & Turk, 2003). Given that worry is prominent in most
anxiety disorders (and depression), we suggest the PSWQ be included when assessing
general anxiety. A copy of the PSWQ can be found in Molina and Borkovec (1994) or
Appendix B of Antony et al. (2001).
Daily Mood Rating
In clinical practice daily idiographic ratings of general anxiety level can be a very useful
metric for tracking fluctuations in subjective anxiety. For example, Craske and Barlow
(2006) suggest that individuals complete a Daily Mood Record in which overall anxiety,
maximum anxiety, overall physical tension, and preoccupation with worry are rated on
a 0 (none) to 100 (extreme) scale at the end of each day. This can be augmented with
single ratings on more specific symptom dimensions that may be more indicative of the
person’s particular anxiety disorder such as ratings on average worry about having a
panic attack in panic disorder or mean daily social evaluative anxiety in social phobia.
It is important that the cognitive therapist also assess changes in general anxiety as
part of an evaluation of treatment effectiveness and for identifying situations that trigger anxiety. These data can be useful for suggesting issues that need to be addressed in
therapy. We have found a 0 to 100 single scale most helpful in capturing the day-to-day
changes in general anxiety (see Figure 5.2).
0
50
100
“Absolutely no anxiety,
totally relaxed”
“Moderate or usual level
of anxiety felt when in
anxious state”
“Extreme, panic-stricken
state that is unbearable
and feels life-threatening”
FIGURE 5.2. Daily mood rating scale.
Cognitive Assessment and Case Formulation
135
This rating scale has been incorporated into a daily situation record form (see
Appendix 5.1) that can be used to assess daily fluctuations in general anxiety.
Beck Depression Inventory–II
The Beck Depression Inventory–II (BDI-II; Beck, Steer, & Brown, 1996) is a 21-item
questionnaire that assesses the severity of cognitive-affective, behavioral, and somatic
symptoms of depression over a 2-week interval. The BDI-II is the third and latest revision
of the original BDI that was published by Beck, Ward, Mendelson, Mock, and Erbaugh
(1961). The second revision of the BDI (Beck & Steer, 1993) has been used widely in
depression research and so most of the psychometric information has been generated on
that measure. However, the BDI and BDI-II are highly correlated (r = .93; Dozois, Dobson, & Ahnberg, 1998), so the psychometric findings on the BDI are relevant for BDIII. Although the BDI appears to be mulifactorial, the Total Score is most often used in
clinical practice and research (Beck, Steer, & Garbin, 1988). There is extensive research
demonstrating the internal reliability and the convergent and discriminant validity of
the BDI (see Beck et al., 1988, for review; Tanaka-Matsumi & Kameoka, 1986). Individuals with major depression score significantly higher (M = 26.52, SD = 12.15) than
those with anxiety disorders (M = 19.38; SD = 11.46; see Beck et al., 1996). The cutoff scores for the BDI-II are 0–13 nondepressed, 14–19 mildly depressed or dysphoric,
20–28 moderately depressed, and 29–63 severely depressed (Beck et al., 1996; see also
Dozois et al., 1998). Given the high co-occurrence of depressive symptoms and disorder
in those with high anxiety, it is recommended that the BDI-II be included in the standard assessment battery for anxiety. The BDI-II is available from Pearson Assessment
at pearsonassess.com.
Clinician Guideline 5.2
To assess severity of general anxiety symptoms, administer the BAI, CCL, PSWQ, and daily
ratings of average anxiety level. If desired, DASS Anxiety can be included and the HRSA can
be used when clients over- or underreport their level of anxiety. The BDI-II should be added
to assess level of comorbid depressive symptoms. A complete assessment will also include
measures of specific anxiety disorders that are reviewed in subsequent chapters.
FEAR ACTIVATION: ASSESSMENT AND FORMULATION
Based on the cognitive model (see Figure 2.1) in this section we focus on assessment tools
that provide critical information needed to develop a case formulation of the immediate fear response and its consequences. Experimental research on the immediate fear
response uses information-processing tasks and psychophysiological measures that are
not readily available to the therapist. However, the practitioner can use self-report, interview, and behavioral observation methods that rely on conscious, effortful processing in
a manner that offers valuable information on a person’s immediate fear response.
One of the most basic propositions of cognitive therapy is that schematic content,
which is inaccessible to direct observation or detection, can be inferred from conscious,
136
ASSESSMENT AND INTERVENTION STRATEGIES
verbal report of one’s thoughts, images, daydreams, ruminations, evaluations, and the
like. Beck (1967) stated: “The schemas pattern the stream of associations and ruminations as well as the cognitive responses to external stimuli. Hence, the notion of schemas
is utilized to account for the repetitive themes in free associations, daydreams, ruminations, and dreams, as well as in the immediate reactions to environmental events”
(p. 283). If schemas direct conscious thought, then the differential activation and content of schemas can be inferred from verbal content (see also Kendall & Ingram, 1989).
Furthermore, there is a direct link between automatic and elaborative processes as indicated by evidence that changes in conscious appraisal or meaning can modify automatic
threat biases (see Mansell, 2000) and that an automatic attentional bias can be induced
through an attentional training program that involves both brief and long processing
intervals (e.g., Matthews & MacLeod, 2002; MacLeod et al., 2002). Together these
considerations lead to the following proposition: that the nature and function of automatic threat schema activation during the initial fear response can be determined from
the cognitive, behavioral, and physiological products of this activation.
Three primary questions must be addressed in any case formulation of the immediate fear response (Phase I).
•฀ What situations, cues, or experiences trigger the immediate fear response?
•฀ What is the core schematic threat or danger to self?
•฀ What is the immediate inhibitory or defensive response to this threat?
Although standardized questionnaire and interview data can be helpful in building a
case formulation, the most critical information will be obtained from idiographic measures. These are self-monitoring forms, rating scales, and diary records that allow the
person to collect critical information when experiencing anxiety. They are tailored to
the particular needs and circumstance of each client so that process-oriented, “online”
data gathering is available that contributes to a more accurate case conceptualization.
Behavioral observation is another assessment approach that can provide important
clinical information on immediate fear response. Some anxiety states like social phobia, OCD, and PTSD can be quite easily elicited in the therapy session by introducing
relevant triggers of anxiety. Other anxiety disorders like panic and GAD require more
ingenuity in order to trigger an immediate fear response. Often the therapist accompanies the client to particular external situations in order to observe an anxious state. In
either case, direct observation of a fear response provides opportunity to gather detailed
information on the nature, severity, and functional characteristics of the immediate fear
response. We believe it is important for the therapist to have at least one opportunity to
observe a client’s acute anxiety state in order to develop an accurate case formulation
and a sensitive individually tailored treatment plan.
Clinician Guideline 5.3
Daily self-monitoring and direct behavioral observation are important assessment strategies that should be a regular feature of any assessment and case formulation of anxiety. Both
strategies are critical for determining the nature of immediate fear activation.
Cognitive Assessment and Case Formulation
137
Situational Analysis
A cognitive case conceptualization of anxiety must begin with a thorough assessment of
the situations, experiences, and cues that trigger anxiety. The cognitive therapist could
begin at the most general level by asking about the problems or difficulties that led to
a decision to seek treatment. With the anxiety disorders, the development of a Problem
List (see Persons & Davidson, 2001) will inevitably lead into a discussion of the situations that trigger anxiety. Three types of situations should be assessed (see also Antony
& Rowa, 2005). Table 5.1 presents a number of clinical questions that can be asked in
the assessment interview.
Environmental Triggers
Information on the external or internal cues, situations, or experiences that trigger a
state of fear or anxiety is a critical part of an evidence-based assessment strategy for the
anxiety disorders (Antony & Rowa, 2005). It is important that the cognitive therapist
obtain a comprehensive list of anxiety-provoking situations with sufficient detail to fully
understand the specific cues that trigger an anxious response. In practically all cases,
objects, events, or situations in the external environment can be identified that trigger
anxiety. Examples of anxiety-eliciting situations include a variety of social settings or
interactions in social phobia, in GAD daily events involving some degree of uncertainty
or possibility of negative outcome (e.g., going on a trip, scheduling an appointment,
paying bills), or in OCD situations that elicit fear of contamination or doubt would be
prominent (e.g., washroom, sitting on park bench). Since a comprehensive knowledge of
anxiety-eliciting situations is critical to case formulation, treatment planning, and later
exposure interventions, the therapist should complete a broad list of triggering situations that range from the mild to most severe anxiety-arousing triggers.
The cognitive therapist can obtain initial information on environmental triggers
from the clinical interview by asking specific questions about the types of situations that
elicit anxiety (see Table 5.1). However, most anxious clients have selective and inaccurate recall of their anxiety-provoking situations so daily self-recording forms should be
assigned in the early phase of treatment. Appendix 5.2 provides a Situational Analysis
Form that can be used to collect key information on provoking situations. In some cases
where there has been a long history of avoidance or where the client’s self-report may be
unreliable, it may be necessary to interview a spouse, close friend, or family member to
obtain more complete information on provoking situations. The therapist could accompany the client to particular situations or set a homework assignment that involved
exposure to a situation in question in order to assess its anxiety- eliciting properties.
However, this might be too threatening for many anxious individuals, especially in the
early phase of treatment.
Interoceptive Triggers
Most anxious individuals have a heightened awareness and responsiveness to the bodily
sensations that characterize physiological hyperarousal in anxiety. Physiological sensations such as increased heart rate, feeling warm, lightheadedness, weakness, tension,
and the like can themselves become triggers for elevated anxiousness. Thus it is impor-
138
ASSESSMENT AND INTERVENTION STRATEGIES
TABLE 5.1. Interview Questions for Assessing Different Types of Situational Triggers in Anxiety
Type of situational
triggers
Clinical questions
External situations,
settings, objects
•฀ “Have you noticed whether there are certain situations or experiences that are
most likely to cause you to become anxious?”
•฀ “Are there some situations that cause only mild anxiety or occasionally cause
you to be anxious and other situations that cause more extreme levels of
anxiety?”
•฀ “Can you tell me about the last time you were in each of these situations and felt
anxious.” [Therapist probes for a full account of anxiety-provoking situations by
obtaining examples from client’s immediate past.]
•฀ “Have you noticed whether there is anything about a situation that might make
the anxiety worse?”
•฀ “Is there anything about a situation that might ease your anxiety?”
•฀ “How often do you experience these situations in your daily life?”
•฀ “Do you try to avoid the situation? How much does this interfere in your daily
life?”
Interoceptive
(physical) cues
•฀ “When you are in an anxious situation, have you noticed any changes in how
you feel physically?” [Therapist could mention a few of the most common signs
of hyperarousal if client needs prompting.]
•฀ “Have you noticed whether any of these physical sensations occur before you
start feeling anxious?”
•฀ “How often do you get these physical sensations when you’re anxious? Are some
always present whereas others are only present occasionally?”
•฀ “Which of the physical sensations is felt most strongly when you are anxious?
Which of the sensations do you notice first when you’re anxious?”
•฀ “Have you noticed whether you feel more anxious once you are aware of a
physical sensation?” [e.g., client might feel more anxious about sudden increases
in heart rate.]
•฀ “Have you ever had the physical sensation (e.g., chest pain) occur unexpectedly
when you were not anxious? Can you recall an example of when this happened?
How did you feel after noticing the sensation?”
•฀ “Do you take special precautions to ensure that you don’t experience a
particular physical sensation?” [e.g., client might avoid time pressures because
wants to maintain state of calm and avoid feeling tense.]
Cognitive cues
•฀ “Have you ever had a thought, image, or impulse about something quite
weird, unexpected, even disturbing, suddenly pop into your mind?” [Therapist
might have to give examples or provide client with a list of common unwanted
intrusions to prompt self-report of intrusions.]
•฀ “When you are entering an anxious situation [therapist states specific situations],
do you recall having any sudden thoughts or images pop into your mind?”
•฀ “Do any of these unexpected intrusive thoughts involve things that are totally
out of your character or that would cause you considerable embarrassment or
dreaded consequences?”
•฀ “How upset do these thoughts make you feel?”
•฀ “Have you ever felt concerned that something might be wrong with you or
that something bad could happen because of the intrusive thought, image, or
impulse?”
Cognitive Assessment and Case Formulation
139
tant to determine if there are any particular bodily sensations that make clients feel
more anxious. Although interoceptive cues to anxiety are particularly evident in panic,
they will be present in all of the anxiety disorders (Antony & Rowa, 2005). For example, a person with social phobia might become even more anxious in a social setting if
she begins to feel warm because this is interpreted as a sign of increased anxiety that
might be noticed by others.
The therapist should include questions in the clinical interview about interoceptive
cues (see Table 5.1), but many clients have even less insight into the presence of physical
triggers to anxiety than they do to external cues. A self-monitoring checklist of physical
sensations, such as the form in Appendix 5.3, can be assigned as homework in order
to gather more accurate information on interoceptive triggers. An interoceptive exposure test is another useful strategy for assessing the physical triggers of anxiety. Taylor
(2000) describes a number of exercises that can be used in the therapy session to induce
physical sensations. For example, the client can be asked to breathe through a straw or
jog on the spot to induce chest tightness, to tense muscles to induce trembling/shaking,
or to face a heater to feel bodily sensations of warmth. Although the intentional induction of such sensations can not be equated with the spontaneous occurrence of these
sensations in vivo, they give the therapist an opportunity to directly observe the client’s
reaction to the sensations.
Cognitive Triggers
Unwanted and disturbing intrusive thoughts, images, or impulses are an example of a
cognition that can trigger anxiety. Practically everyone experiences unwanted mental
intrusions and they are commonly found in all the anxiety disorders. First described by
Rachman (1981) within the context of OCD, unwanted intrusive thoughts, images, or
impulses are “any distinct, identifiable cognitive event that is unwanted, unintended,
and recurrent. It interrupts the flow of thought, interferes in task performance, is associated with negative affect, and is difficult to control” (Clark & Rhyno, 2005, p. 4). Some
examples of common intrusions are “unprovoked doubt about locking the door when I
know I did,” “touching something gross and dirty that is lying on the street,” “saying an
insulting or embarrassing remark for no apparent reason,” “blurting out an obscenity in
a public meeting,” “swerving your car into oncoming traffic,” and the like. Unwanted
intrusions are very common in OCD as obsessions and in PTSD as sudden recollections
of a past trauma. However, they can also occur in GAD as a negative consequence of
excessive worry (Wells, 2005a) or as unwanted cognitions in the presleep phase of individuals suffering from insomnia (Harvey, 2005). Unwanted intrusions often involve the
theme of losing control that leads to a dreaded negative consequence.
It is important that the cognitive therapist inquire about unwanted intrusive
thoughts. Table 5.1 lists some possible questions for assessing this clinical phenomenon.
With the exception of OCD or PTSD, individuals are often not very aware of their
intrusive thoughts. A list of common unwanted intrusions can be used and clients asked
if they ever had any of these thoughts, images, or impulses (lists can be found in D. A.
Clark, 2004; Rachman & de Silva, 1978; Steketee & Barlow, 2002). Since most intrusions are provoked by external cues, clients can be asked to be especially vigilant for
mental intrusions when in situations that typify their anxious concerns.
140
ASSESSMENT AND INTERVENTION STRATEGIES
Elements of a Complete Situational Analysis
A thorough situational analysis should consist of the following elements:
•฀ Detailed description of multiple situations or triggers
•฀ Intensity of associated anxiety
•฀ Frequency and duration of exposure to situation/trigger
•฀ Presence of escape, avoidance responses
•฀ Specific eliciting cues
A detailed description of each situation or trigger is needed. Subtle changes in context can alter the intensity of anxiety. For example, a patient with panic disorder might
report little anxiety driving to work on a very familiar route. However, vary the route
by one new street, and anxiety level might change dramatically. The proximity of a
safety signal will also influence anxiety (e.g., presence of a trusted friend or distance
from a medical facility). It may be that a particular situation (e.g., interacting with
work colleagues) needs to be broken down into finer gradients in order to understand
its anxiety-eliciting properties. The cognitive therapist should have enough detail about
each anxiety-provoking situation or trigger so that accurate exposure assignments can
be constructed.
It is important to know the intensity of anxiety felt in each situation since the therapist should have a range of situations or triggers that elicit mild to severe anxiety states.
Some clients require considerable practice using the 0–100 rating scale to gauge their
anxiety level, especially if they tend to engage in dichotomous thinking (e.g., they feel
intensely anxious or not anxious at all). These ratings, however, are needed to develop
an effective treatment plan.
The therapist must determine how often the person experiences an anxietyprovoking situation and the duration of his or her exposure to the situation. Anxietyprovoking situations that occur regularly in the person’s daily life will be more helpful
for treatment than rare or exceptional occasions. For example, daily social interactions
with work colleagues that trigger anxiety in someone with social phobia will be much
more important to treatment than a situation such as giving a speech that may rarely
occur in the person’s life. Also, does the provoking situation involve brief or prolonged
exposure when the person encounters the circumstance? Again anxiety-provoking situations that involve longer exposure intervals (e.g., using a public washroom) will be
more useful in treatment planning than triggers involving brief exposure (e.g., touching
a public telephone as you walk by it).
The cognitive therapist should also obtain information on the extent that each situation is associated with escape or avoidance. Clients should be asked if they always try
to avoid the situation or escape from the situation as quickly as possible. At this assessment stage the therapist should have a good understanding of how well the client tolerates anxiety in each provoking situation. If the situation is avoided on some occasions
but not others, what determines the presence or absence of avoidance? Does this depend
on the person’s mood state or some subtle characteristic of the situation? Information on
escape and avoidance will be critical in planning an exposure hierarchy.
Finally, the cognitive therapist should determine whether there are specific cues or
stimuli in a situation that are first noticed by the anxious individual. For example, when
Cognitive Assessment and Case Formulation
141
a person with fear of contamination first enters a public area, what is first noticed that
elicits some concern, the speck of dirt on the floor or the fact that a stranger just brushed
past him? For a socially anxious individual, does he first notice that his throat feels dry or
that his hand seems to be shaking? A person with PTSD might avoid a particular route to
work because of anxiety but it is really the fact that driving past a particular ethnic store
along the route triggers flashbacks that is the crux of the problem. In addition it is important to determine whether the person is hyperviligant for these subtle and specific anxietyprovoking cues. It is likely that a self-monitoring homework assignment will be needed in
order to identify the salient attentional features of anxiety-provoking situations.
Clinician Guideline 5.4
A complete situational analysis should include detailed information on a wide range of
anxiety-provoking external and internal situations or stimuli, with a specific focus on the
intensity of anxiety, frequency, and duration of situational exposure, extent of escape/avoidance, and presence of eliciting cues.
The First Apprehensive Thoughts or Images
One of the main consequences of threat schema activation during the immediate fear
response phase is the production of automatic threat-oriented thoughts and images (see
Figure 2.1). These threat-oriented automatic thoughts and images occur at the earliest
point in anxiety generation and provide a window into the schematic content that is the
basis of the anxiety disorder.
In the context of assessment the cognitive therapist can refer to these initial threatoriented automatic thoughts as the first apprehensive thoughts. They are defined as brief,
sudden, and completely automatic thoughts or images that something bad or unpleasant is about to happen, or at least could happen, to persons or their valued resources.
In panic disorder these first apprehensive thoughts might refer to the dangers posed by
a perceived physical sensation, in social phobia it might be the thought of drawing the
attention of others, in OCD it might be of some catastrophe for others as a result of
one’s action or inaction, in PTSD it could be a sense of losing control and increased personal vulnerability, and in GAD it could be the occurrence of some serious negative life
event. Notice that the first apprehensive thoughts always reflect some important aspect
of the person’s primary anxious concerns. In fact the situational analysis will provide
the therapist some clues as to first apprehensive thoughts because of the types of situations that provoke anxiety.
Discovering the client’s first apprehensive thoughts presents special challenges for
assessment. Often these thoughts are so rapid and transient that the person only experiences them as a sudden feeling of fear or apprehension. The actual automatic thought
content is lost because it is quickly replaced by more elaborative, reasoned reappraisal
of the situation. So, when the therapist questions clients about their first apprehensive
thoughts, what is recalled and reported are the more deliberate reappraisal thoughts
that occur in the second phase of anxiety. Clients interviewed when not feeling anxious
may dismiss the first apprehensive thoughts as too exaggerated and unrealistic, and so
deny they ever occurred during an anxious episode.
142
ASSESSMENT AND INTERVENTION STRATEGIES
So how can the cognitive therapist gain access to this fleeting cognitive content?
It is important to introduce the topic of first apprehensive thoughts in a collaborative,
exploratory manner. A description of the first apprehensive thoughts should be provided
and the client should be warned that it is often difficult to identify these thoughts in
the anxiety cycle. Explain that when entering an anxious situation, most people are so
focused on how they feel and the details of the situation, that their first apprehensive
thoughts are often lost to them. The following example can be used to introduce clients
to the concept of the first apprehensive thoughts.
“Imagine for a moment that you are walking down a deserted street or country
road by yourself and it is getting dark. Suddenly you hear a noise behind you. You
immediately stiffen, your heart beats quickly, and you quicken your pace. Why this
sudden surge of adrenalin? No doubt you instantly interpret the noise as a dangerous possibility: ‘Could someone be approaching from behind who could cause me
harm?’ You turn around and there is no one there. Quickly you think to yourself
‘No one is there, it must have been the wind, a squirrel, or my imagination.’ It is
this secondary thought, this reevaluation of the situation, that sticks in your mind.
If later I asked you about your walk, you would remember a momentary twinge
of fear and the later realization that ‘nothing was there.’ That first apprehensive
thought that triggered the fear ‘Is there an attacker behind me?’ is lost to recall,
instead replaced by your reasoned response to the situation.
“In the last couple of sessions you have described a number of situations that
cause you considerable anxiety. In these situations you would have had some initial
apprehensive thoughts or images that fueled your fear or anxiety. It may be that
now you can’t remember what they are because you don’t feel threatened at the
moment and you are not in an anxiety-provoking situation. However, it is important to our treatment that we discover the first apprehensive thoughts. We want to
know what ‘kick-starts’ the anxiety. Together, by carefully going over each situation and collecting some further information, we may discover the types of apprehensive thoughts or images that define your anxious experiences.”
The first assessment strategy for identifying the initial apprehensive thoughts is the
clinical interview. Although individuals often don’t remember their initial automatic
anxious thoughts, a few specific, well-phrased questions can provide some initial clues
to these thoughts. Here are some examples of clinical questions:
•฀ “You indicated that in situation X you feel intensely anxious. For you what would
be the worst thing that could happen in this situation? What would be the worst
possible outcome? Try to think about the worst consequence without considering
whether or not you think it is likely to happen.”
•฀ “Is there anything specific about the situation or about how you are feeling that
concerns you? What is not quite right for you? What is different from your normal self?”
•฀ “How could the situation change so you feel less concern, less uneasy?”
•฀ “What do you tell yourself to ease your anxiety, to reassure yourself that everything will be fine?”
Cognitive Assessment and Case Formulation
143
It is important that the first apprehensive thought be recorded in the person’s own words
and not reflect the therapist’s own suggestions. The therapist might probe for a certain type of thought content, but its actual expression should reflect the idiosyncratic
concerns of the client. This will ensure that the apprehensive thought content is highly
relevant to the specific anxious concerns of the client.
It is also important to remember that even in the same anxiety-provoking situation, people will differ in the focus of their apprehension and so it is important for the
therapist to discover each client’s unique anxious apprehension. As an example, a client
reports intense anxiety about going to a meeting with work colleagues. The first apprehensive thought could be any of the following possibilities:
•฀ “What if I’m asked a question in the meeting that I can’t answer? Everyone will
think I’m incompetent.” (performance evaluation cognition)
•฀ “What if I have to say something and everyone stares at me? This makes me so
nervous.” (social evaluation cognition)
•฀ “What if my voice trembles when I speak? Everyone will know that I am nervous
and wonder what’s wrong with me.” (social phobia cognition)
•฀ “What if I have a panic attack in the meeting?” (panic disorder cognition)
•฀ “What if I accidentally blurt out an insulting remark?” (OCD cognition)
•฀ “What if I’m not supposed to be at this meeting and everyone wonders why I am
there?” (interpersonal acceptance cognition)
•฀ “What if I feel nauseated in the meeting and have to run out and vomit?” (cognition about specific fear of vomiting)
•฀ “I never really know what to say in these meetings and how to chit-chat with others; I really hate this.” (social skills deficit cognition)
As can be seen from this example, there are a large number of possible apprehensive
thoughts triggered by any anxiety-provoking situation. The purpose of the cognitive
assessment is to identify the anxious thought content that is unique to each client.
Self- monitoring homework tasks must be assigned in order to obtain more immediate and accurate assessment of the first apprehensive thoughts or images. The “immediate anxious thoughts” column of the Situational Analysis Form (Appendix 5.2) can
be used as an initial attempt to collect self-monitoring data on the first apprehensive
thought. Clients must be encouraged to focus on “what is the worst that could happen
in this situation” without considering whether it is probable, realistic, or rational. They
should be encouraged to write down the automatic threat thoughts while they are in
the anxious situation. They can ask themselves “What is so bad about this situation?”,
“What am I thinking is the worst that could happen?,” or “What could harm me in
this situation?”. If a more detailed self-monitoring form is needed, the Apprehensive
Thoughts Self-Monitoring Form can be used (see Appendix 5.4).
Imagery or role plays can be used in the therapy session to determine individuals’ apprehensive cognitions in anxious situations. In fact anxious patients often have
conscious fantasies or images of physical or psychosocial harm that can elicit intense
subjective feelings of anxiety (Beck et al., 1974). It is important, then, that the therapist
determine whether the initial apprehension may take the form of an intrusive image
such as reliving a traumatic event. Whatever the case, the client can be asked to imagine
144
ASSESSMENT AND INTERVENTION STRATEGIES
a recent anxiety-provoking situation or the therapist and client could role-play the situation in order to elicit automatic anxious thoughts or images. Throughout, the therapist
probes for a client’s anxious appraisals of the situation and her ability to cope. Naturally
the effectiveness of this assessment approach depends on the client’s imaginative ability
or capacity to engage in role playing.
Induction exercises can also be used to elicit apprehensive thoughts. For example,
various physiological hyperarousal symptoms can be induced and clients encouraged
to verbalize their “stream of thoughts” as they experience these symptoms. A situation
could be created in the therapy session or stimuli introduced to elicit anxiety and clients
could again be asked to verbalize their emerging thoughts. For example, someone with
fear of contamination could be given a dirty cloth to touch and then report on his anxious thoughts.
Finally, the most effective procedure for eliciting the first apprehensive thoughts is
to accompany the client into a naturalistic anxiety- provoking situation. Although the
presence of the therapist might have a safety cue effect, careful probing of clients’ stream
of consciousness should reveal their first apprehensive thoughts. Even generating an
expectation of exposure to an anxiety-provoking situation might be sufficient to elicit
these primary automatic anxious thoughts.
Clinician Guideline 5.5
Obtain an accurate assessment of the client’s first apprehensive thoughts in a variety of
anxiety-provoking situations to determine the underlying threat schema responsible for the
anxious state.
Perceived Autonomic Arousal
Individuals are usually very aware of the physical symptoms of anxiety and so can
quite readily report these symptoms in the clinical interview. They should be asked for
examples of recent anxiety episodes and the physical symptoms experienced at these
times. Rather than have clients report on the typical anxiety attack, it is better for
them to report on specific incidents of anxiety and the exact physical symptoms experienced during these episodes. Some variation in the physical symptoms of anxiety can be
expected across different anxiety episodes.
The practitioner will be relying mainly on clients’ self-report of their physiological
responses since use of psychophysiological laboratory-based or ambulatory equipment
for monitoring purposes is rarely feasible in the clinical setting. Self-monitoring forms
should be used for clients to collect “online” data of their physiological responses when
anxious. In most cases the Physical Sensation Self-Monitoring Form (Appendix 5.3) can
be given as a homework assignment and will provide the needed information on the
client’s autonomic arousal profile. In certain cases where physiological arousal plays a
particularly important role in the persistence of anxiety (i.e., panic disorder, hypochondriasis), an expanded checklist of bodily sensations can be administered (see Appendix
5.5).
Three questions must be addressed when assessing subjective physiological hyperarousal in the immediate fear response phase. First, what is the typical physiological
Cognitive Assessment and Case Formulation
145
response profile when the person is in a state of heightened anxiety? It is important to
determine whether the client typically experiences the same physiological symptoms in
a variety of anxiety-provoking situations. Which bodily sensations are most intense?
Which arousal symptoms are experienced first? How long do they persist? Does the
person do anything to achieve relief from the hyperarousal?
A second question concerns how the state of physiological hyperarousal is interpreted. Are there certain bodily sensations that are the primary focus of attention? What
is the client’s concern or fears about that sensation? Identifying the exaggerated threat
appraisal of a particular body sensation is another important source of information on
the core threat schemas that are driving the anxiety. Table 5.2 presents the exaggerated
threat appraisals and schemas that may be associated with a number of physiological
hyperarousal symptoms.
A final question when assessing physiological arousal is their role in the persistence
of anxiety. Catastrophic misinterpretation of physical symptoms plays a key role in
panic disorder (D. M. Clark, 1986a) and hypochondriasis (Salkovskis & Bass, 1997)
but may be less prominent in OCD or GAD. In anxiety disorders where misinterpretation of physical symptoms is a prominent concern, treatment will focus on “decatastrophizing” these exaggerated appraisals. Thus case formulations for anxiety must take
into account the nature, interpretation, and function of physiological hyperarousal during the phase of immediate fear.
Clinician Guideline 5.6
The nature, function, and interpretation of physiological hyperarousal and other bodily sensations must be determined as part of any case formulation for anxiety.
TABLE 5.2. Exaggerated Threat Appraisals and Schemas That May Be Associated with Common
Physical Symptoms of Anxiety
Physical sensation
Exaggerated faulty appraisal
Threat-oriented schema
Difficulty breathing,
shortness of breath
“I can’t breath properly, I feel like I’m
not getting enough air.”
Risk of slow, agonizing death by
suffocation
Chest tightness, pain,
heart palpitations
“Maybe I am having a heart attack.”
Death from sudden cardiac arrest
Restless, agitated
“I am losing control; I can’t stand this
feeling of anxiety.”
Risk of going crazy, embarrassing
myself, being overwhelmed with
unending anxiety, etc.
Dizzy, lightheaded,
faint
“I might be losing consciousness.”
Might never regain consciousness;
cause embarrassment by fainting in
public
Nausea
“I might be sick to my stomach;
vomit.”
Suffocate from vomiting;
embarrassment from being sick in
public setting
Note. Based on Taylor (2000).
146
ASSESSMENT AND INTERVENTION STRATEGIES
Immediate Inhibitory Responses
Immediate, defensive responses such as escape, avoidance, freezing, or fainting (Beck et
al., 1985, 2005) are part of an automatic inhibitory strategy to reduce fear. An important part of any cognitive assessment of anxiety is to identify these fear-inhibiting
responses and yet their detection can be difficult because they are so automatic, with
the individual having little conscious awareness of their presence. However, it is important to determine the presence of these responses because they should be targeted for
change given their capacity to reinforce the anxious state and undermine treatment
effectiveness. As an example, a number of years ago one of us treated a woman with
driving fear after having been rear-ended in a motor vehicle accident. Upon assessment
it was discovered that while in traffic she anxiously kept her eye on the rearview mirror,
checking to ensure that the car behind her was not too close. This checking behavior
was done quite automatically as a defensive response. However, it meant that she was
not attending as closely as she should to the traffic in front of her, thus increasing the
likelihood of another accident.
Once again a detailed clinical interview, self-monitoring, and behavioral observation during heightened anxiety are the primary assessment approaches for identifying
immediate defensive behaviors. There are a number of subtle defensive reactions that the
clinician should be aware could occur as an immediate inhibitory response.
•฀ Avoids eye contact to threatening stimulus (e.g., socially anxious person fails to
make eye contact when conversing with others).
•฀ Cognitive avoidance in which attention is shifted away from a disturbing thought
or image (e.g., in PTSD a trauma-related intrusion might trigger a state of dissociation).
•฀ Immediate escape (flight) behavior (e.g., a person with fear of contamination
quickens her pace as she walks past a park bench where homeless people sit).
•฀ Behavioral avoidance (e.g., a person with mild agoraphobia automatically chooses
a less crowded store aisle).
•฀ Reassurance seeking (e.g., a person keeps reciting the phrase “There is nothing
to fear”).
•฀ Compulsive response (e.g., a person automatically pulls the car door handle
repeatedly to make sure it is locked).
•฀ Defensive physiological reflex response (e.g., a person anxious about swallowing
food starts to gag when attempting to swallow; a person with driving fear stiffens
body or is generally tense whenever he is a passenger in a car).
•฀ Tonic immobility (freezing) (e.g., during a brutal assault a person may feel paralyzed, feeling like she is unable to move [see Barlow, 2002]).
•฀ Fainting (e.g., a person experiences a sudden drop in heart rate and blood pressure at the sight of human blood or mutilated bodies).
•฀ Automatic safety behaviors (e.g., a person automatically clutches an object to
avoid falling or losing balance).
Given the automatic, rapid nature of these defensive responses, it is likely that some
form of behavioral observation will be necessary to accurately assess their presence. It
would be preferable if the cognitive therapist accompanied the client into anxious situ-
Cognitive Assessment and Case Formulation
147
ations and then noted any inhibitory responses. Alternatively, a friend, family member,
or spouse could be given the above list of defensive responses and asked to note whether
any of these responses were observed when accompanying the client in anxious situations.
Clinician Guideline 5.7
Uncover automatic cognitive and behavioral inhibitory responses through behavioral observation to identify reactions that could later undermine the effectiveness of exposure.
Cognitive Processing Errors
Cognitive processing during the immediate fear response tends to be highly selective,
with attention narrowly focused on the source of threat and one’s ability (or inability)
to deal with this threat. As a result certain unintended errors will be evident in the client’s evaluation of the threat that will not be readily apparent to the individual. These
cognitive errors can be determined from the automatic anxious thoughts and behaviors
that are elicited in anxiety-provoking situations. Appendix 5.6 provides a list of the common cognitive errors seen in anxiety disorders, followed by a self-monitoring form that
clients can use to become more aware of their anxious processing biases. This should be
introduced after the client has been taught how to identify the first apprehensive thought.
Teaching clients how to identify their cognitive errors will not only provide information
for the case formulation but it is a useful cognitive intervention strategy (see Chapter 6).
Many anxious clients have difficulty identifying the cognitive errors in their anxious thinking. It may take a number of sessions before the client can capture examples
of his own thinking biases. In the meantime the therapist can use the form in Appendix
5.6 to identify some of the thinking errors that are apparent from the clinical interview
and self-monitoring of anxious thoughts. This can be incorporated in the case formulation until more accurate data are available from clients’ own recording of their thinking
errors.
Clinician Guideline 5.8
Use Appendix 5.6, Common Errors and Biases in Anxiety, to train clients to identify the
automatic cognitive errors that occur whenever their anxiety is provoked by certain internal
or external triggers.
SECONDARY REAPPRAISAL: ASSESSMENT AND FORMULATION
Anxiety is always the result of a two-stage process involving the initial activation of
threat followed by a slower, more reflective processing of the threat in light of one’s coping resources. For this reason the cognitive therapist also assesses secondary elaborative
processing, focusing on two questions that must be addressed in the case conceptualization.
148
ASSESSMENT AND INTERVENTION STRATEGIES
1. How does the individual’s more elaborative reappraisal of the situation lead to
an increase in anxiety?
2. How effective is the individual’s reflective reappraisal in reducing or terminating
the anxiety program?
Assessment of secondary reappraisal is not as difficult as assessment of the immediate fear response because these processes are less automatic and so more amenable
to conscious awareness. Individuals tend to have more insight into these slower, more
deliberate processes that are responsible for the persistence of anxiety. Because cognitive therapy tends to focus on this secondary level, an accurate assessment of elaborative
processes is critical to the success of the intervention. In this section we examine five
domains of secondary processing that should be included in the assessment.
Evaluation of Coping Abilities
Reliance on maladaptive coping strategies and failure to adopt healthier responses to
threat are considered key factors in failed emotional processing in general and the persistence of anxiety in particular (e.g., Beck et al., 1985, 2005; Wells, 2000). One of
the most common distinctions in the coping literature is between strategies that focus
on emotion regulation versus those that focus directly on life problems. Lazarus and
Folkman (1984) originally defined emotion-focused coping as “directed at regulating
emotional response to the problem” (p. 150) and problem-focused coping as “directed
at managing or altering the problem causing the distress” (p. 150). There is now a large
body of research indicating that certain aspects of emotion-focused coping (e.g., rumination) are related to the persistence of negative emotional states, whereas problemfocused coping is associated with reduction in negative affect and the promotion of positive emotion and well-being (e.g., Carver, Scheier, & Weintraub, 1989; see reviews by
Fields & Prinz, 1997; Folkman & Moskowitz, 2004; for discussion of positive aspects
of emotion expression, see Austenfeld & Stanton, 2004 ).
In the present context this distinction between an emotion- and a problem-focused
approach is useful in understanding the persistence of anxiety. Coping responses that
focus on “how can I make myself feel less anxious” are more self-defeating (i.e., lead
to persistence of unwanted anxiousness), whereas coping that is more problem-oriented
(i.e., “I have a real life problem that I must address” ) is more likely to lead to a reduction in anxiety.
The cognitive therapist should keep this distinction in mind when assessing the coping responses of anxious clients. To what extent is the client’s coping repertoire dominated by emotion-focused versus problem-oriented strategies? In addition three other
questions on coping must be addressed in the assessment:
1. How often does an individual use various maladaptive and adaptive coping
responses when feeling anxious?
2. What is the client’s perception on the effectiveness of the coping strategies in
reducing anxiety?
3. Does the client perceive that an increase or persistence of anxiety is associated
with the coping response?
Cognitive Assessment and Case Formulation
149
Appendix 5.7 provides a checklist of 34 behavioral and emotional coping responses
that pertain to anxiety. We suggest the therapist go over the checklist as part of the clinical interview since most clients should be quite aware of their coping responses when
anxious. Also most anxious individuals probably have not considered the perceived
effectiveness of their coping and its effects on the intensity and duration of anxiety.
Therefore some probing and questioning may be necessary in order to obtain this information.
From this assessment one should be able to specify in the case formulation which
maladaptive coping strategies are frequently associated with anxiety and their perceived
effectiveness, the relative effectiveness of any adaptive strategies that the client already
employs, and the overall level of confidence or helplessness felt in dealing with anxiety.
This will also provide the therapist with clues about behavioral changes that may be
targeted in treatment. However, it is also likely that this checklist assessment must be
complemented with questions about coping responses that may be unique to the specific
anxiety disorders. Also many of the strategies listed in Appenditx 5.7 could be stress
management responses. Therefore it is important that clients be asked to focus on activities employed directly in response to their anxiety and not activities they use to relieve
general stress, improve mood state, or enhance their overall sense of well-being.
Clinician Guideline 5.9
Use Appendix 5.7, Behavioral Responses to Anxiety Checklist, to assess how often various
behavioral and emotional coping strategies are used to control anxiety. Highlight the role of
these strategies in the persistence of anxiety in the case conceptualization.
Deliberate Safety-Seeking Behavior
White and Barlow (2002) define safety behaviors as “those actions that a patient engages
in to help him or her feel more secure or protected” (p. 343). The focus of safety behaviors is to feel secure, safe, which has the obvious benefit of reducing feelings of anxiety
(see Chapter 3, Hypotheses 2 and 7, for further discussion).
It is important to clearly identify in the case formulation the main safety-seeking
responses whether they are more automatic and habitual in nature or more consciously
mediated, deliberate coping responses. By this point in the assessment much of this information has already been collected from individuals’ self-monitoring of their responses in anxious situations (i.e., Situational Analysis Form, Apprehensive Thoughts Self-Monitoring
Form) or from the previous evaluation of coping strategies (i.e., Behavioral Responses to
Anxiety Checklist). The cognitive therapist can go back over these forms and select out
responses that often occur when the person is anxious. For each response the following
questions should be asked to assess the safety-seeking function of the response:
•฀ “I notice from your form that you often do X [state actual response] when you
feel anxious. To what extent do you feel safer or more secure after you have done
this? [e.g., How much safer do you feel going to the mall with a friend versus
going to the mall alone?]”
150
ASSESSMENT AND INTERVENTION STRATEGIES
•฀ “What would happen to your anxiety if you did not engage in this safety activity?
[e.g., What would happen to your anxiety if you didn’t carry your medication
with you?]”
•฀ “How important is this activity to your way of dealing with or managing your
anxiety? Is it something you do deliberately or is it more automatic, like a habit
that you are hardly aware of doing?”
Once the client’s primary safety-seeking responses have been identified, it is important to also specify the cognitions and physical sensations associated with safety seeking
(i.e., Salkovskis, Clark, et al., 1999). This might be quite obvious from the cognitivebehavioral responses recorded on the self-monitoring forms or on occasion the cognitive
therapist might have to assess more specifically. The following fictitious clinical excerpt
illustrates the type of inquiry that could be used to identify safety-seeking cognitions.
THERAPIST: I notice from the checklist that you indicated you always carry your Ativan
with you at all times. Could you tell me why this is so important for you?
CLIENT: Well, I just feel better knowing that I have the medication if I ever need it. I
haven’t used the Ativan in months but knowing that it is there makes me feel better.
THERAPIST: What would happen if you forgot to take the medication bottle with you?
CLIENT: I know that I would feel a lot more anxious if I realized I didn’t have it. The
Ativan is so effective in relieving my anxiety. If I have it with me, I know that I
could always take a pill if the anxiety gets too severe. Even though I haven’t used the
medication in months, just knowing that the anxiety can’t get out of hand because
I could always take an Ativan seems to help.
THERAPIST: Is there anything you feel or experience when in an anxious situation that is
somehow better just knowing you have the medication?
CLIENT: Well as you know I get really afraid of having another panic attack when I
notice that I’m becoming more anxious. The worst thing is feeling like I am losing
control. Knowing that I could take an Ativan and be calmer and in control within a
few minutes makes me feel a lot better; it makes me feel more confident.
A number of cognitions are evidently associated with this client’s medication-related
safety-seeking behavior. She believes just having access to the medication gives her more
confidence and makes her feel safer, more secure. More importantly, there is a direct
functional relationship between the catastrophic thought “of losing control” and being
able to take the medication. This belief that the medication is an important source of
regaining control and thwarting overwhelming anxiety will become a target in treatment. If the cognitive basis of safety seeking cannot be determined by interview or
review of the self-monitoring forms, direct observation of the client’s anxiety either by
accompanying the person into an anxious situation or conducting an anxiety-induction
exercise in the session might be necessary. In all anxiety cases identifying the primary
safety-seeking behaviors and their cognitive basis is an important part of the case formulation for anxiety.
Cognitive Assessment and Case Formulation
151
Clinician Guideline 5.10
Identify the primary intentional safety-seeking behaviors by reviewing the client’s Behavioral
Responses to Anxiety Checklist (Appendix 5.7) and determine the functional significance
and cognitive basis of the responses. Also reconsider the safety-seeking function that may be
associated with the more automatic, inhibitory reactions noted in Guideline 5.7. This should
result in a clear specification of the subtle, more automatic and more conscious, deliberate
safety-seeking behaviors that characterize the client’s anxiety.
Constructive Mode
An important part of the secondary phase of anxiety is the activation of a more constructive, problem-oriented approach to the threatening situation. It must be recognized that all treatment-seeking individuals will have some capacity to respond to their
anxiety in a more constructive manner. It is important to identify these strengths in the
case formulation so this can be incorporated into the treatment plan. What behavioral
responses to anxiety does the client already exhibit that indicates a more constructive
approach? Is the person able to engage in adaptive problem-solving? Are there any cognitive strategies that lead to a reduction in the perceived level of threat? It is useful to
assess the constructive mode when the person is in a nonanxious state. How do they
perceive threat and their personal vulnerability when not anxious? How well can they
bring this more realistic, adaptive perspective to bear when they are anxious? How difficult is it to believe the constructive perspective when anxious?
Very often individuals who seek cognitive therapy for anxiety have had previous
treatment or read cognitively oriented self-help books on anxiety. Thus it is very likely
that some constructive response to their anxiety is already present. Table 5.3 presents
various types of constructive responses to anxiety and sample clinical questions that can
be used to assess constructive mode activation when anxious.
Assessment of clients’ “spontaneous” use of various constructive approaches to
anxiety is important for two reasons. First, it provides some indication of the clients’
strengths around which a treatment plan can be formulated. And second, it may be that
a particular constructive approach has not been employed effectively and so the client
has negative expectations about its success. It would be important for the therapist to
know this before assigning this strategy as homework. In sum, assessment of constructive mode activation is an important part of the case formulation.
Clinician Guideline 5.11
Identify adaptive coping strategies that are present in the client’s repertoire and the extent to
which these responses are utilized during periods of anxiety. Evaluation of the constructive
mode should also include an assessment of the client’s ability to engage in a more realistic
appraisal of his or her anxious concerns when not anxious and whether this more realistic
perspective is available during anxious episodes.
152
ASSESSMENT AND INTERVENTION STRATEGIES
TABLE 5.3. Examples of Constructive Responses to Anxiety That Should Be Assessed as Part
of the Case Conceptualization
Constructive response
Clinical questions
Spontaneous exposure
•฀ How often does the client deliberately expose himself to anxiety-provoking
situations?
•฀ How intense and for how long is the anxiety tolerated before escape occurs?
•฀ Does exposure occur on a regular basis? Are safety cues present or absent?
•฀ What is the client’s evaluation of the exposure experience? Is it seen as
reducing her anxiety or exacerbating it?
Self-initiated response
prevention
•฀ How often does the client inhibit responses that are intended to reduce
anxiety (e.g., a compulsive ritual in OCD)?
•฀ How hard is it to resist the urge to engage in the anxiety-reduction activity?
•฀ Does resistance occur on a regular basis?
•฀ How is the attempt to resist the anxiety-reducing activity evaluated? Is the
resistance viewed as making the anxiety worse or better?
Relaxation response
•฀ How often does the client engage in progressive muscle relaxation, controlled
breathing, or meditation in response to anxiety?
•฀ What is the client’s evaluation of the effectiveness of these strategies in
managing anxiety?
•฀ Is there any evidence that the client is using relaxation as an escape strategy
because of a fear of being anxious? To what extent is relaxation an adaptive
or maladaptive response strategy for anxiety?
Problem-solving
ability
•฀ Does the client take a problem-solving approach to the source of anxiety?
(e.g., a student worried about exam failure works on improving study skills)
•฀ What is the perceived effect of these problem-solving attempts on anxiety
level?
•฀ Are there any weaknesses to the problem-solving strategy that may undermine
its positive effect on anxiety?
Realistic threat
reappraisal
•฀ Does the client engage in any questioning or reappraisal of his initial threat
appraisal, and if so, how effective is this questioning?
•฀ Can he practice evidence gathering where he seeks out contrary information
that the threat is not as great as initially thought?
•฀ Does he ever turn to some form of empirical hypothesis testing where he seeks
out experiences to determine if his fears are realistic or exaggerated?
Reappraisal of
personal vulnerability
•฀ Does the client engage in any form of evidence gathering about her ability to
cope with the threat?
•฀ Can she recall past experiences of successful coping as a means of readjusting
her initial sense of personal vulnerability?
•฀ Does she deliberately engage in anxiety-provoking activities to test-out her
vulnerability?
Cognitive Coping and the Role of Worry
Excessive Worry
We previously argued that worry in highly anxious individuals is an important contributor to the persistence of anxiety because of domination of threat mode activation
(Beck & Clark, 1997; see Chapter 2). It is a detrimental cognitive coping strategy (see
Chapter 3, Hypothesis 10) that is evident in most of the anxiety disorders, especially
GAD. Thus it is important that the nature, extent, and function of worry be assessed
when developing a case formulation for anxiety.
Cognitive Assessment and Case Formulation
153
The first question to address is whether the client worries when anxious and, if so,
what is the worry content, its frequency, and its persistence. The therapist can expect
worry content to broadly fit within the main anxious concerns of the client. For example, in panic disorder the worry is about disturbing bodily sensations, whereas in social
phobia worry about performance in social settings and the evaluation of others are
dominant.
Appendix 5.8 presents the Worry Self-Monitoring Form A that can be used to assess
any worry content associated with anxious episodes. This can be given as a homework
assignment or the cognitive therapist could complete the form in the therapy session
based on anxious situations identified on the Situational Analysis Form or the Apprehensive Thoughts Self-Monitoring Form. The purpose of the Worry Self-Monitoring Form A
is to collect qualitative information on any worry themes that may play an important role
in the persistence of the worry. This worry content will provide useful information for
cognitive interventions the therapist will employ later in treatment. It is also important
to determine how often the client worries when anxious and the duration of the worry
episode. Worry that is frequent and lasts for 1–2 hours has a very different treatment
implication from the occasional bout of worry that is dismissed within a few minutes.
In Chapter 3 we discussed a number of negative consequences associated with worry
that may account for its pathological effects on anxiety (e.g., heightened sensitivity to
threat information, increased sense of personal vulnerability, an increase in unwanted
intrusive thoughts, an escalation in negative emotions, cognitive/emotional avoidance,
and ineffective problem solving). However, most individuals will not have sufficient
insight into the negative effects of worry to allow collection of this information from a
homework assignment. Instead the therapist could use the worry episodes recorded on
the Worry Self-Monitoring Form A as the basis for questioning that explores the negative consequences of worry. The following is a therapy excerpt based on a client with
social phobia who was anxious about interrupting his supervisor to ask an important
question:
THERAPIST: John, I notice from the Worry Self-Monitoring Form that you were particularly anxious on Friday about having to go to your supervisor’s office to ask an
important question about a project you were trying to finish. You rated your anxiety as 80/100 and the first apprehensive thought was “he is going to be so angry
that I am interrupting him with such a stupid question.”
JOHN: Yeah, I was really upset about this situation. These kinds of things really bother
me. I find I get so anxious.
THERAPIST: It appears that you spent approximately a half hour worrying about this
before you went and then you were worried most of the day afterward that your
supervisor was angry with you for interrupting him. You’ve written that before you
asked the question you were worried mainly about his angry reaction (i.e., would he
be abrupt with me), whether you would be able to make yourself clearly understood,
and whether you would understand your supervisor’s response. Afterward you kept
replaying the conversation in your mind to determine whether you sounded stupid or not. Furthermore, you worried about your supervisor’s opinion of you and
whether this would reflect negatively on your year-end performance evaluation.
You’ve also written that you were worried that others overheard the conversation in
154
ASSESSMENT AND INTERVENTION STRATEGIES
your supervisors’ office and were thinking that you were “so pathetic” (using your
expression).
JOHN: I find I worry a lot about how I come across to other people and the negative
effects of my “bumbling” conversations with others.
THERAPIST: John, in this situation did you notice any changes in your anxiety level while
you were worrying before or after the interaction with your boss?
JOHN: Not sure what you mean.
THERAPIST: Did you notice any increases or decreases in your anxious feelings while you
were worrying?
JOHN: Oh, I definitely felt more anxious. Before the interaction I tried to convince myself
everything would be okay but all I could think about was his anger, and afterward I
again tried to reassure myself that everything would be fine but the more I thought
about it the more convinced I became that he thinks I’m incompetent.
THERAPIST: So one of the negative effects of worry is that it makes you more anxious
rather than less anxious. Do you think worrying about talking to your boss made
you more effective when you actually went and asked him the question?
JOHN: No, I don’t think worrying about it gave me more confidence or improved the
conversation. All I could think about was getting it over and dealing with the negative consequences later.
THERAPIST: You’ve mentioned a couple of other ways that worry may have a negative
effect. It sounds like it makes you think about avoiding or escaping as quickly as
possible. Also it doesn’t sound like worry helps you cope with situations or problems more effectively. Did you notice anything else about your thinking when you
were worrying?
JOHN: Not sure what you mean.
THERAPIST: Did you notice whether a lot of upsetting thoughts kept popping into your
mind even though you didn’t want them?
JOHN: Oh, yes. I kept seeing an image of my supervisor’s angry face, I could hear him
shouting at me, and I kept having the thought “He thinks I’m such an idiot.”
THERAPIST: From your description, John, it sounds like worry has a number of negative
effects on your anxiety. It is associated with an increase in anxious feelings; it may
interfere with your ability to deal with situations; it intensifies the urge to escape or
avoid the anxiety; and it increases unwanted distressing thoughts and images. This
is not unusual in anxiety. Our research on worry indicates that it has far-reaching
negative effects that can contribute to the persistence of anxiety. Would you like to
make worry reduction an important goal in your anxiety treatment plan?
JOHN: Yeah, I definitely think I need to learn how to get a handle on my worry.
Other Cognitive Coping Strategies
In Chapter 3 (see Hypothesis 10) attempts to deliberately suppress unwanted thoughts
and feelings were considered compensatory coping strategies that may contribute to the
persistence of anxiety. In addition the intentional suppression of emotional expression
may have adverse effects on negative emotion, although far fewer studies have investi-
Cognitive Assessment and Case Formulation
155
gated this possibility. An assessment of intentional thought suppression and emotional
inhibition should be included in the case formulation. Appendix 5.9 presents a cognitive
coping checklist that includes emotion inhibition along with a number of other intentional thought control strategies that may exacerbate the anxious state.
The Cognitive Responses to Anxiety Checklist (Appendix 5.9) can be assigned as
a homework exercise. However, most anxious clients are probably not aware of their
thought control strategies because these responses can become quite habitual over time.
Thus some training and education will be required to teach clients how they may engage
in maladaptive thought control strategies that only make anxious thoughts more salient.
One might be able to review a recent anxious episode and use the checklist to determine
which of the 10 strategies occurred and to what extent they contributed to anxiety
reduction. Alternatively, a state of anxiety could be induced in the therapy session (or
observed in a naturalistic setting) and clients could be asked whether they use any of the
checklist strategies to control their anxious thoughts or worries.
Another way to highlight the nature of thought control in clients’ experience of
anxiety is to conduct a modified thought suppression experiment. This is illustrated in
the following example.
THERAPIST: Lorraine, I would like to take a closer look at your anxiety about having a
panic attack. You’ve indicated that often you feel your chest tighten and your first
apprehensive thoughts are “I must be getting anxious. I really need to calm down.
If I don’t I could have another of those terrible panic attacks.”
LORRAINE: Yes, that is exactly how I feel. I really hate those feelings and would do anything to get rid of them.
THERAPIST: Okay, what I would like to do is a little exercise with you right here in the
office. First, I would like to see if you could focus on your anxious thoughts right
now. Maybe you could bring these thoughts to your mind by tightening your chest
muscles or imagining you are in a recent anxious situation. It doesn’t matter how
you do it, but I would like you to think about feeling anxious and the possibility of
having a panic attack.
LORRAINE: I’m not sure I want to do this. I’m afraid I could trigger a panic attack.
Already I’m beginning to feel anxious.
THERAPIST: I understand your concern. We can stop the exercise at any time. I simply
want you to bring the anxious thoughts to your full attention. If you are beginning
to feel anxious, then maybe you can focus on these anxious thoughts right now
even without tightening your chest muscles.
LORRAINE: Oh, I have no trouble thinking about my anxiety right now and the possibility of a panic attack.
THERAPIST: Okay, Lorraine, please close your eyes and focus your attention on thoughts
of being anxious right now. Think about how you are feeling and the last time you
had a panic attack. I am going to ask you to hold that thought for 30 seconds. . . .
[pause] Now stop thinking about your anxiety. I am going to give you another 30
seconds to stop thinking about your anxiety and the possibility of panic. You can
do this anyway you choose. . . . Okay, stop [pause]. Were you able to stop yourself
from thinking about your anxiety and the possibility of a panic attack?
156
ASSESSMENT AND INTERVENTION STRATEGIES
LORRAINE: This is really hard. I tried not to have the thoughts but it was almost
impossible. I think it was too short. I needed more time to get rid of my anxious
thoughts.
THERAPIST: It is true that I gave you only half a minute. However, many people find the
exercise even more frustrating if I drag it out longer. The important point is whether
or not you were able to stop the anxious thinking.
LORRAINE: Not really. I seemed to be getting more and more anxious the harder I tried
to get the thoughts out of my mind.
THERAPIST: You’ve just made an important point. The harder you try “not to think
about the anxiety, the more you think about it.” I have here a checklist of various
strategies people use to change their anxious thinking. [Therapist passes Lorraine
a copy of the Cognitive Responses to Anxiety Checklist.] Could you look through
this checklist and tell me whether you just used any of these strategies in your
attempt to not think about the anxiety.
LORRAINE: Well, I tried to deliberately not think about the anxiety (item #1), and I kept
telling myself it is stupid to be anxious because I’m sitting here in your office (#6),
and I tried to convince myself that I couldn’t possibly have a panic attack right now
(item #3). None of this seemed to work very well, though.
THERAPIST: From this exercise we’ve discovered a couple of things. First, you’ve reported
that the harder you try to control your anxious thoughts, the worse they get. And
second, you’ve reported a number of different mental control strategies you used to
try and get rid of anxious thoughts. I realize that you’ve just done a “simulation”
because in real life your anxious thoughts and feelings would be a lot more intense
than they were while you were sitting in this office. I wonder how often you might
automatically try to control your anxious thoughts whenever you feel anxious using
the same strategies you reported just now. And I wonder what effect this might have
on your anxiety. I wonder if it makes your anxiety worse or better. Would you like
to find out?
LORRAINE: Sure, I think that would be a good idea.
THERAPIST: Okay, before our next session, could you take a copy of the Cognitive
Responses to Anxiety Checklist that we just used and see if you could capture some
times when you were anxious. Try to focus on your attempts to control the anxious
thinking. Which of these thought control strategies did you use and how effective
were they? Under the “how often” category, just check whether you used the strategy or not. You don’t have to capture all your anxious moments, just one or two a
day. It should take only a few minutes each day to fill out the form. Do you think
that is doable?
LORRAINE: Yeah, I should be able to do this in the next week. I’m still having lots of
anxiety.
Clinician’s Guidelines 5.12
Assessment of the nature, frequency, and function of worry and other cognitive control
responses is an important aspect of the case formulation of the persistence of anxiety. The
Worry Self-Monitoring Form A (Appendix 5.8) can be used to obtain clinical information
Cognitive Assessment and Case Formulation
157
on worry, and the Cognitive Responses to Anxiety Checklist (Appendix 5.9) is available to
assess deliberate thought control strategies.
Threat Reappraisal
This final aspect of case conceptualization is a culmination of all the assessment activities that have been described previously. As clients consciously and deliberately reflect
back on their anxiety when in a safe and relaxed context, what is their evaluation of
the threat and their ability to cope? Appendix 5.10, the Anxious Reappraisal Form,
can be used to explore with clients their threat and vulnerability cognitions when feeling anxious and then their evaluation of the threat and personal vulnerability when
calm, not anxious. One would expect that when anxious the thinking should be biased
toward exaggerated threat and underestimated ability to cope, whereas during periods
of no anxiety the person’s threat evaluation would be more realistic and self-confidence
elevated.
The Anxious Appraisal Form should be used as a clinical resource in the therapy
session to help the therapist explore and then record the client’s anxious and nonanxious
appraisals rather than assigned as a homework exercise. The cognitive therapist should
point out the differences between the client’s thinking when anxious and not anxious.
It should be emphasized that the client is capable of thinking in a more realistic fashion
about her anxious concerns when in a calm and relaxed state. This means that the goal
of therapy is to help clients learn to generalize their more realistic thinking about the
threat and their ability to cope to their most difficult anxious moments. In this way the
information obtained on the Anxious Reappraisal Form can be used to define one of the
primary treatment goals of cognitive therapy for anxiety.
Clinician Guideline 5.13
Use the Anxious Reappraisal Form (Appendix 5.10) to assess clients’ ability to generate a
more realistic reappraisal of threat and personal vulnerability during periods of no anxiety.
This can be used to highlight the biased, exaggerated nature of their thinking when anxious.
Shifting to the more realistic appraisal that is evident in low anxiety should be a stated goal
of treatment.
CASE FORMULATION OF ANXIETY: A CASE ILLUSTRATION
Cognitive Case Formulation
We conclude this chapter with a case illustration to demonstrate how the clinician can
utilize the theory-driven assessment perspective described in this chapter to arrive at
an overall cognitive case conceptualization of anxiety. Although we have described
a very detailed cognitive approach to assessment and case formulation, it should be
obvious from the following case presentation that much of the critical information can
be obtained from the clinical interview, self-monitoring forms, observation of anxiety
within the session, and standardized diagnostic interview and questionnaire measures.
Thus it is reasonable to expect that an initial cognitive case conceptualization can be
158
ASSESSMENT AND INTERVENTION STRATEGIES
developed within the first two to three sessions, which will then be frequently revised
and elaborated throughout the treatment process. In fact it is this changing, evolving
nature that is the heart of case conceptualization (Persons, 1989).
A diagram of the cognitive case conceptualization of anxiety that is available in
Appendix 5.11 can be used to summarize the assessment information and derive an
individualized case formulation. Although there are many components to case formulation, the clinician is never expected to have a “finalized formulation” before initiating
treatment. Certain core elements of the conceptualization should be apparent after the
initial assessment and prior to treatment such as the situational triggers, first apprehensive (automatic anxious) thoughts, physiological hyperarousal, defensive (i.e., safetyseeking) responses, primary worry content (if relevant), and coping strategies. These
aspects of the formulation will be revised and other components completed during subsequent treatment sessions. An individualized case formulation, then, evolves over the
course of therapy.
Cognitive Case Conceptualization
We return to the clinical case presented at the beginning of this chapter. Sharon sought
treatment for a long-standing problem with persistent anxiety that manifested itself
mainly while interacting with work colleagues in her employment setting.
Diagnostic and Symptom Assessment
Sharon was administered the ADIS-IV as well as the general anxiety measures discussed
in this chapter. Based on the ADIS-IV her primary axis I disorder was social phobia.
Panic disorder without agoraphobic avoidance was a secondary axis I diagnosis. She
also met criteria for a past major depression, single episode. The depression spontaneously remitted after 2 months and occurred in response to the death of a pet. She also
reported a subclinical fear of heights and worry, but the latter was clearly related to her
social anxieties at work. She obtained the following scores on the questionnaire battery;
Beck Anxiety Inventory Total = 6, Beck Depression Inventory–II Total = 12, Hamilton Anxiety Rating Scale = 10, Cognitions Checklist— Depression = 15 and Cognitions Checklist—Anxiety = 7, and Penn State Worry Total = 64. Sharon also completed
the Social Phobia and Anxiety Inventory (SPAI; Turner, Beidel, & Dancu, 1996) and
obtained a Difference Score of 105.9, which is consistent with untreated generalized
social phobia. Thus the psychometric data suggest only mild anxiety symptoms that are
more cognitive than physiological in nature. The Penn State Worry score is elevated, but
this is due to the client’s worry about her social interactions at work. The BDI-II and
CCL-D suggest the presence of some depressive symptoms. A pretreatment average daily
anxiety level of 21/100 again confirmed a rather low level of anxiety.
The diagnostic assessment clearly indicated that the social phobia should be the
primary focus of treatment. Although she met diagnostic criteria for panic disorder,
the initial onset was 15 months ago, with the last full-blown panic attack occurring 1
year ago. In total she experienced four full-blown panic attacks and a number of limited
symptom attacks, with many of the later occurring in social contexts at work. However,
Sharon reported only minimal, brief periods of concern about the panic attacks that
lasted only 3–4 days after a full-blown episode. Sharon also indicated that the panic
Cognitive Assessment and Case Formulation
159
attacks had limited interference in her daily functioning. Thus it was concluded that
treatment of panic attacks that were not related to her social anxiety was not warranted
at this time.
Assessment of Immediate Fear Response
Sharon listed a number of situations that trigger her anxiety at work. These include
speaking up or interacting in a small group meeting, talking to persons in authority
like her supervisor, one-to-one interaction with work colleagues over their computer
problems, and initiating phone calls at work. These activities were associated with moderate to severe anxiety and a moderate level of avoidance. Given that her job primarily
involves consultation with others, Sharon was frequently confronted with these anxietyprovoking situations on a daily basis. Other social activities that triggered considerable anxiety and avoidance were going to parties and being assertive, especially refusing unreasonable requests. Sharon completed a Situational Analysis Form as part of a
homework assignment and reported a number of anxious episodes focused on small
meetings and one-to-one interaction at work. The only cognitive trigger to anxiety was
the anticipatory thought “I must speak to my supervisor about this problem.” It was
decided to target her anxiety in small meetings and one-to-one interaction with work
colleagues since these represented the main triggers to her anxiety.
Two main automatic apprehensive thoughts became apparent from Sharon’s selfmonitoring homework assignments and subsequent interview sessions. When anticipating or first encountering a social situation at work, Sharon would think “I hope I am
able to perform okay” and “I hope my face doesn’t turn red.” The only physiological
sensations she reported when anxious was feeling warm and her face turning red (i.e.,
blushing). Blushing was a major concern for Sharon. She interpreted this as a sign that
she was anxious, losing concentration, and would be less able to speak clearly and sensibly to others. She was also concerned that people would notice that her face was red
and wonder what was wrong with her.
As a result of these anxious cognitions and the negative interpretations of blushing,
Sharon exhibited a number of automatic defensive responses. Behaviorally she would
say as little as possible in meetings (i.e., avoidance) and would speak very rapidly when
she was forced to interact with others (i.e., escape response). She avoided eye contact
in her social interactions. She also was hypervigilant about feeling warm and would
often touch her face or check in a mirror to determine if she was visibly red. Her main
automatic cognitive defense was to reassure herself that everything was okay and to try
to relax. In sum her primary automatic defensive response to ensure safety was to say
as little as possible in social situations, to avoid eye contact, and to locate herself in a
setting so as to draw as little attention as possible.
A number of cognitive errors were evident in Sharon’s anxious thinking about social
situations. Catastrophizing was apparent in her belief that having a red face was highly
abnormal and something that others would also interpret as a sign of abnormality. She
was also convinced that once her face turned red, it meant she was anxious and would
lose her concentration. This would result in poor performance, which others would
evaluate as social incompetence. Tunnel vision was another cognitive error since Sharon
would often become preoccupied with her face and whether she was feeling warm in
social settings. She also engaged in emotional reasoning in that feeling uncomfortable
160
ASSESSMENT AND INTERVENTION STRATEGIES
in social settings meant that she was in greater danger of not functioning well and more
likely to draw the attention of others. Finally, she tended to think of anxiety from an allor-nothing perspective with certain situations associated with social threat and so intolerable, whereas other situations were entirely safe (e.g., working alone in her office).
Assessment of Secondary Reappraisal
Sharon exhibited a number of deliberate coping strategies in response to her social anxiety. She would try to physically relax in social situations by engaging in deep, controlled
breathing, she tried to answer questions via e-mail in order to avoid face-to-face interaction with work colleagues, she would procrastinate about such things as asking her
supervisor for clarification on an issue, and she was quiet and withdrawn in meetings,
saying as little as possible. She also tried to suppress her feelings to hide any sense of discomfort. The intentional use of alternative means of communication with others (e.g.,
e-mail) had a prominent safety-seeking function. These strategies were all somewhat
effective in reducing her social anxiety. Sharon was concerned that if she changed her
approach to social anxiety if might make her work life more stressful.
Worry played a secondary role in Sharon’s social anxiety. She worried on a daily
basis about the possible social interactions she might encounter, whether she would
experience a lot of anxiety throughout the day, and whether she would be socially
incompetent as a result. She also worried outside the work setting that the extra stress
and anxiety she was feeling at work might have a negative effect on her health and wellbeing. Sharon’s cognitive coping strategies to control her anxiety were quite limited
other than the use of reassurance and rationalization that everything will be fine and
self-instructions to control her anxiousness. She concluded she was generally ineffective
in controlling the anxiety and that the best strategy was to minimize social contact as
much as possible. Interestingly, this perspective on social threat and vulnerability was
evident even when she was not anxious and alone.
Treatment Goals
Based on our cognitive case conceptualization, the following goals were developed in
Sharon’s treatment plan:
•฀ Decatastrophize her misinterpretation and maladaptive beliefs about blushing
and the consequent negative evaluation of others.
•฀ Modify the belief that anxiousness in social settings must be controlled because
it will lead to dire negative outcomes such as social incompetence (i.e., reappraise
the probability and severity of threat).
•฀ Reduce avoidance and increase exposure to socially anxious situations.
•฀ Eliminate maladaptive defensive and coping strategies such as speaking too quickly
when anxious, reliance on deep breathing, and self-rationalization focused on
convincing herself there is no threat.
•฀ Reduce the negative effects of worry about being anxious whenever social interaction is anticipated.
•฀ Improve assertiveness and other verbal communication skills when interacting
with authority figures such as a supervisor.
Cognitive Assessment and Case Formulation
161
SUMMARY AND CONCLUSION
In this chapter we presented a cognitive case conceptualization perspective that is based
on the cognitive model of anxiety (see Chapter 2). Although this framework will be
applicable to all anxiety cases, it will require some modification for each of the specific
anxiety disorders. Case formulation plays an important role in cognitive therapy for
all psychological problems. For the anxiety disorders assessment begins with clinical
diagnosis and administration of standardized questionnaires. It is important that presence of anxious and depressive symptoms be assessed. Utilizing interview methodology,
self-monitoring forms, and direct observation, the clinician gathers information on the
immediate or automatic cognitive, physiological, and behavioral responses that characterize the initial fear program. This is followed by assessment of more deliberate cognitive and behavioral coping strategies that are intended to terminate the anxious episode
but instead inadvertently contribute to its long-term persistence. Particular attention
is given to automatic and intentional responses that have a safety-seeking function.
The assessment will culminate in a specification of the threat and personal vulnerability appraisals generated when the individual is an anxious and a nonanxious state.
This detailed cognitive formulation should lead to the development of specific treatment
goals that will guide the intervention process. A Quick Reference Summary is provided
in Appendix 5.12 to assist the clinician in applying our cognitive perspective on assessment and case formulation in clinical practice.
APPENDIX 5.1
Daily Anxiety Ratings and Situation Record
Name:
Date:
Instructions: Use the rating scale below to record a number from 0 to 100 that indicates the
average level of anxiety you experienced during the day. In the far right column briefly describe any
situations that you found particularly anxiety provoking on a particular day.
0
50
100
“Absolutely no anxiety,
totally relaxed”
“Moderate or usual level
of anxiety felt when in
anxious state”
“Extreme, panic-stricken
state that is unbearable
and feels life-threatening”
Day of the
Week/Date
Rating of Average
Anxiety Level
(0–100)
Provoking Situations
(Note any situations that increased your anxiety
during the day)
1. Sunday
2. Monday
3. Tuesday
4. Wednesday
5. Thursday
6. Friday
7. Saturday
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
162
APPENDIX 5.2
Situational Analysis Form
Name:
Date:
Directions: Please write down any situations that triggered an anxiety response. Very briefly describe the situation in column two and in the
third column rate the intensity of anxiety (0–100) and its duration (number of minutes). In the fourth column note the most prominent anxious
symptoms you experienced and in the fifth column record any immediate thoughts in the situation. In the final column please comment on your
immediate response to the anxiety.
Date/Time
Situation
Anxiety Intensity
(0–100) and Duration
(min)
Primary Anxious
Symptoms
Immediate Anxious
Thoughts
Immediate Response to
Feeling Anxious
1.
163
2.
3.
4.
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright 2010 by The Guilford Press. Permission to photocopy
this appendix is granted to purchasers of this book for personal use only (see copyright page for details).
APPENDIX 5.3
Physical Sensation Self-Monitoring Form
Name:
Date:
Directions: Please write down any situations or experiences that caused an increase in your anxiety.
Pay particular attention to whether you experienced any of the bodily sensations listed on this form
while you were in that situation. Use the rating scales beside each sensation to indicate how you
felt about the bodily reaction.
1. Briefly describe anxious situation:
Record level of anxiety in situation (0–100 scale):
Checklist of physical sensations experienced in situation:
Physical Sensation
Intensity of Physical Sensation
Anxiousness about
Physical Sensation
[Use 0–100 scale defined below]
[Use 0–100 scale defined below]
Chest tightness
Elevated heart rate
Trembling, shaking
Difficulty breathing
Muscle tension
Nausea
Lightheaded, faint, dizzy
Weak, unsteady
Feeling warm, sweaty
Dry mouth
(cont.)
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
164
APPENDIX 5.3 (page 2 of 2)
2. Briefly describe anxious situation:
Record level of anxiety in situation (0–100 scale):
Checklist of physical sensations experienced in situation:
Physical Sensation
Intensity of Physical Sensation
Anxiousness about
Physical Sensation
[Use 0–100 scale defined below]
[Use 0–100 scale defined below]
Chest tightness
Elevated heart rate
Trembling, shaking
Difficulty breathing
Muscle tension
Nausea
Lightheaded, faint, dizzy
Weak, unsteady
Feeling warm, sweaty
Dry mouth
Rating Scale Instructions: Intensity of Physical Sensations Scale, 0 = barely felt the sensation;
50 = strong sense of the sensation; 100 = dominant, overwhelming feeling.
Anxiousness about Physical Sensations Scale, 0 = not at all anxious about having the sensation;
50 = considerable concern that I am having this sensation; 100 = feel intensely anxious, panicky
that I am having this sensation.
165
APPENDIX 5.4
Apprehensive Thoughts Self-Monitoring Form
Name:
Date:
Directions: Please write down any situations or experiences that caused an increase in your anxiety. After rating the level of anxiety experienced
in the situation in the second column, write down your response to the questions posed in the next columns based on what you were thinking
and feeling in the situation. Try to fill in this form while you are in the anxious situation or as soon afterward as possible.
What’s not right about the
situation? [What’s disconcerting
Anxiety-Provoking Situation
Average
[Describe briefly in a few words and Anxiety Level
include date and time of day]
[0–100 scale]
Worst Possible Outcome
[What’s the worst possible thing
that could happen regardless of
how unlikely or unrealistic?]
about the situation or about how
you feel or could behave? Or how
could others behave toward you
that would be upsetting?]
What would ease your
anxiety? [How could the situation
change to ease your anxiety? How
could you change or others change
to ease your anxiety?]
1.
166
2.
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright 2010 by The Guilford Press. Permission to photocopy
this appendix is granted to purchasers of this book for personal use only (see copyright page for details).
APPENDIX 5.5
Expanded Physical Sensations Checklist
Name:
Date:
Instructions: Below you will find a list of physical sensations that can be experienced during periods
of high anxiety or during panic attacks. Please indicate the intensity of the physical sensation during
a typical anxiety episode or panic attack. The checklist should be completed during the anxiety
episode or as soon afterward as possible. Also please circle the bodily reaction or sensation that
you noticed first during the episode of anxiety.
Physical Sensation
Absent
Slight
Moderate
Severe
Very
Severe
Tense muscles
Muscle pain
Weakness
Muscle twitches, spasms
Numbness in hands, feet (or
pins and needles sensation)
Tingling in hands, feet
Nausea
Stomach cramps
Indigestion
Feeling of urgency to urinate
Diarrhea
Congested, buildup of mucus
in throat or nose
Dry mouth
(cont.)
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
167
APPENDIX 5.5 (page 2 of 2)
Physical Sensation
Absent
Slight
Difficulty taking a deep breath,
shortness of breath
Throat feels constricted (like
you could choke)
Chest tightness
Chest pain
Heart pounding, palpitations
Heart skipping a beat
Trembling, shaky
Feeling restless, fidgety
Feelings of unreality
Muscle twitches
Dizziness
Feeling lightheaded
Feeling faint
Unsteady, loss of balance
Hot flushes or chills
Sweating
Other sensations (state):
168
Moderate
Severe
Very
Severe
APPENDIX 5.6
Common Errors and Biases in Anxiety
The following is a list of thinking errors that are common when people feel afraid or anxious. You
may find that you make some of these errors when you feel anxious but it is unlikely that you make
all of the errors every time you are anxious. Read through the list of errors with their definition
and examples. Put a check mark beside the ones that are particularly relevant for you. You will
notice the errors overlap because they all deal with different aspects of overestimating threat and
underestimating safety when feeling anxious. After reading through this list, turn to page 170 where
you will find a form that you can use to become more aware of your own thinking errors when
anxious.
Thinking Error
Definition
Examples
Catastrophizing
Focusing on the
worst possible
outcome in an
anxious situation.
•฀ Thinking that chest tightness is sign of a heart attack
•฀ Assuming friends think your comment is stupid
•฀ Thinking you’ll be fired for making a mistake in your report
Jumping to
conclusions
Expecting that a
dreaded outcome
is extremely likely.
•฀ Expecting that you will fail the exam when unsure of a question
•฀ Predicting that your mind will go blank during the speech
•฀ Predicting that you will be extremely anxious if you make the trip
Tunnel vision
Focusing only
on possible
threat-relevant
information while
ignoring evidence
of safety.
•฀ Notice that a person looks bored while you are speaking in a
meeting
•฀ Notice a spot of urine on the floor of an otherwise very clean
public washroom
•฀ Person with combat PTSD experiences flashback when seeing
newsclip of a far-off regional conflict
Nearsightedness
•฀ An individual with OCD is convinced of possible contamination
Tendency to
even coming within a few yards of a homeless person
assume that threat
is imminent (close •฀ Worry-prone individual is convinced he will be fired any day
•฀ Person with fear of vomiting is concerned she is about to
at hand).
become sick to her stomach because she has an “unsettled
feeling”
Emotional
reasoning
Assuming that the •฀ Flying must be dangerous because I feel so anxious when I fly
•฀ Person with panic assumes the likelihood of “losing control” is
more intense the
greater when feeling intense anxiety
anxiety, the greater
•฀ Worry-prone individual is even more convinced something bad
the actual threat.
will happen because she feels anxious
All-or-nothing
thinking
•฀ Person with obsessional doubts is always concerned that the
Threat and safety
light switch is not completely off
are viewed in rigid,
absolute terms as •฀ Person with social anxiety is convinced his work colleagues will
think that he is incompetent if he speaks up
either present or
•฀ Person who experienced past trauma is convinced she must
absent.
avoid anything that reminds her of the past incident
(cont.)
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
169
APPENDIX 5.6 (page 2 of 2)
Identifying Anxious Thinking Errors
Name:
Date:
Instructions: With the handout entitled “Common Errors and Biases in Anxiety” as your reference,
use the form below to write down examples of your own thinking errors that occur when you feel
anxious. Please focus on how you are thinking when you are in anxious situations or anticipating
the situation. Also focus on your most immediate apprehensive thoughts rather than any secondary
reconsideration of the situation.
Thinking Error
Examples of My Own Anxious Thinking Errors
Catastrophizing
Jumping to conclusions
Tunnel vision
Nearsightedness
Emotional reasoning
All-or-nothing thinking
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
170
APPENDIX 5.7
Behavioral Responses to Anxiety Checklist
Name:
Date:
Instructions: You will find below a checklist of various ways that people tend to respond to anxiety.
Please indicate how often you engage in each response when you are anxious, how effectively the
strategy reduces or eliminates anxious feelings, and whether you think the strategy unintentionally
leads to the persistence of your anxiety.
Scale Descriptions: How often do you engage in this response when you feel anxious? [0 = never,
50 = half of the time, 100 = all the time]; When you engage in this response, how effectively
does it reduce your anxiety? [0 = not at all; 50 = moderately effective in reducing anxiety, 100
= completely eliminates my anxiety]; Based on your experience, to what extent do you think this
response contributes to a persistence of your anxiety? [0 = does not contribute at all, 50 = makes
a moderate contribution, 100 = is a very important factor in the persistence of my anxiety]
Effective in
Reducing
How Often
Anxiety
[0–100 scale] [0–100 scale]
Behavioral and Emotional Responses
Increases
Persistence
of Anxiety
[0–100]
1. Try to physically relax (e.g., muscle relaxation,
controlled breathing, etc.)
2. Avoid situations that trigger anxiety
3. Leave situations whenever I feel anxious
4. Take prescription medication
5. Seek reassurance, support from spouse,
family, or friends
6. Engage in a compulsive ritual (e.g., check,
wash, count)
7. Distract myself with activities
8. Suppress my feelings (i.e., hold in my feelings)
9. Use alcohol, marijuana, or other street drugs
10. Get very emotional, tearful
11. Have an anger outburst
12. Become physically aggressive
(cont.)
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
171
APPENDIX 5.7 (page 2 of 2)
Effective in
Reducing
How Often
Anxiety
[0–100 scale] [0–100 scale]
Behavioral and Emotional Responses
13. Speak or act more quickly in a hurried manner
14. Become quiet, withdraw from others
15. Seek medical/professional help (e.g., call
therapist or GP; go to emergency)
16. Use Internet to chat with friend or obtain
information
17. Reduce physical activity level
18. Rest, take a nap
19. Try to find solution to the problem causing me
anxiety
20. Pray, meditate in effort to reduce anxious
feelings
21. Have a smoke
22. Have a cup of coffee
23. Gamble
24. Engage in pleasurable activity
25. Eat comforting food (e.g., favorite junk food)
26. Seek some place that makes me feel safe, not
anxious
27. Listen to relaxing music
28. Watch TV or videos (DVDs)
29. Do something that is relaxing (e.g., take a
warm bath or shower, have a massage)
30. Seek out a person who makes me feel safe,
not anxious
31. Do nothing, just let the anxiety “burn itself
out”
32. Engage in physical exercise (e.g., go to the
gym, run)
33. Read spiritual, religious, or meditative material
(e.g., Bible, poetry, inspirational books)
34. Go shopping (buy things)
172
Increases
Persistence
of Anxiety
[0–100]
APPENDIX 5.8
Worry Self-Monitoring Form A
Name:
Date:
Instructions: Using the form below, please record whether or not you have any worries associated
with your anxiety. In the first column write down some occasions when you are feeling anxious,
then rate the intensity of the anxiety on the 0–100 scale, and then try to capture your first
apprehensive (anxious) thought in the situation. You can go back to the Apprehensive Thoughts SelfMonitoring Form if you need help identifying the apprehensive thought. In the final column write
down anything that worried you about that situation as well as how long the worry lasted (number
of minutes or hours).
Worry Content
[Is there anything that worries you
Intensity
about the situation or the effects
Anxious Situation
of Anxiety
of anxiety? Is there any negative
[Describe briefly and include
[0–100 First Apprehensive consequence that worries you?
date and time]
scale]
(Anxious) Thought
How long did you worry?
1.
2.
3.
4.
5.
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
173
APPENDIX 5.9
Cognitive Responses to Anxiety Checklist
Name:
Date:
Instructions: You will find below a checklist of various ways that people try to control their anxious
and worrisome thoughts. Please indicate how often you engage in each response when you are
anxious and how effective the strategy is in reducing or eliminating anxious thoughts.
Scale Descriptions: How often do you engage in this response when you feel anxious? [0 = never,
50 = half of the time, 100 = all the time]; When you engage in this cognitive strategy, how
effectively does it reduce or eliminate the anxious thoughts? [0 = not at all; 50 = moderately
effective in reducing anxiety, 100 = completely eliminates my anxiety]
Cognitive Control Response to Anxious Thinking
How Often
Strategy
is Used
[0–100 scale]
Effectiveness
in Reducing
Anxious
Thinking
[0–100 scale]
1. Deliberately try not to think about what is making me
anxious or worried.
2. Tell myself that everything will be okay and will turn out
fine.
3. Try to rationalize the anxiety; look for reasons why my
anxious concerns might be unrealistic.
4. Try to distract myself by thinking about something else.
5. Try to replace the anxious thought with a more positive or
comforting thought.
6. Make critical or negative remarks to myself about being
anxious.
7. Tell myself to simply “stop thinking” like this.
8. Think a comforting phrase or prayer.
9. Ruminate on the anxious thought or worry; I keep going
over in my mind what happened in the past or what could
happen in the future.
10. When I start to feel anxious I try to suppress the feelings
so I don’t look nervous or upset.
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
174
APPENDIX 5.10
Anxious Reappraisal Form
Name:
Date:
Instructions: Please complete the form below to record your perspective when feeling anxious and when not feeling anxious. When you are
anxious, describe the worst outcome that you fear most and rate its felt probability from 0 (not at all likely to happen) to 100 (absolutely
expect it to happen). Then indicate how well you think you could cope with the anxiety and rate your level of confidence in yourself from 0 (no
confidence) to 100 (absolute confidence). Next repeat the form when you are not feeling anxious. As you look back on those anxious situations,
what is the expected outcome and what is your perceived ability to cope with your anxiety?
When Feeling Anxious
When Not Feeling Anxious
175
Feared Outcome
Ability to Cope with Anxiety
Expected Outcome
Ability to Cope with Anxiety
[Describe worst outcome and rate
its probability 0–100]
[Describe coping ability and
confidence 0–100]
[Describe most likely outcome and
rate its probability 0–100]
[Describe coping ability and
confidence 0–100]
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright 2010 by The Guilford Press. Permission to photocopy
this appendix is granted to purchasers of this book for personal use only (see copyright page for details).
APPENDIX 5.11
Diagram of Cognitive Case Conceptualization of Anxiety
Name:
Date of Initial Assessment Session:
A. CURRENT DIAGNOSTIC INFORMATION
[Based on ADIS or SCID; duration refers to length of current disorder]
Primary axis I diagnosis
Duration:
Secondary axis I diagnosis:
Duration:
Tertiary axis I diagnosis:
Duration:
Additional subclinical diagnoses:
Number of episodes of primary diagnosis:
B. SYMPTOM PROFILE
Beck Anxiety Inventory Total:
Beck Depression Inventory–II Total:
Cognitions Checklist—Anxiety:
Cognitions Checklist—Depression:
Hamilton Anxiety Rating Scale Total Score (optional):
Penn State Worry Questionnaire Total:
Pretreatment Daily Anxiety Mean (sum of ratings over week/7):
C. PROFILE OF IMMEDIATE FEAR RESPONSE
Situational Analysis
List Primary External Triggers
List Primary Internal/Cognitive Triggers
1.
1.
2.
2.
3.
3.
4.
4.
5.
5.

First Apprehensive Thoughts/Images
List Core Automatic Anxious Thoughts/Images
[present during episodes of anxiety]
1.
2.
3.
4.

(cont.)
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
176
APPENDIX 5.11 (page 2 of 3)
Perceived Physiological Hyperarousal
List Primary Physical Sensations/Symptoms
Misinterpretation of Sensation/Symptom
1.
1.
2.
2.
3.
3.
4.
4.
5.
5.

Automatic Inhibitory/Defensive Responses
List Primary Behavioral Defenses
List Primary Cognitive Defenses
1.
1.
2.
2.
3.
3.
4.
4.
*Mark defenses with safety-seeking function with asterisk.

Primary Cognitive Errors
[evident during anxious episodes]
Type of Cognitive Error
Actual Example of Error from Client Assessment
1.
1.
2.
2.
3.
3.
4.
4.
5.
5.
D. PROFILE OF SECONDARY REAPPRAISAL
Primary Behavioral and Emotional Coping Strategies
Briefly Describe the Coping Strategy
Perceived Effect
in Reducing Anxiety
1.
1.
2.
2.
3.
3.
4.
4.
5.
5.
*Mark coping strategies with safety-seeking function with asterisk.

(cont.)
177
APPENDIX 5.11 (page 3 of 3)
Primary Worry Symptoms
Briefly describe the main worry content during anxious episodes
1.
2.
3.
4.
5.

Principal Thought Control Strategies
Briefly Describe the Control Strategy
Perceived Effect
in Reducing Anxiety
1.
1.
2.
2.
3.
3.
4.
4.
5.
5.
Threat and Vulnerability Appraisal
When Anxious
[briefly summarize the client’s perspective
on threat and vulnerability when anxious]
Threat and Vulnerability Reappraisal
When Not Anxious
[briefly summarize the client’s
perspective on threat and vulnerability
when not anxious]
178
APPENDIX 5.12
Chapter 5 Quick Reference Summary: Cognitive Assessment of Anxiety
I. Conduct Diagnostic Interview (ADIS-IV or SCID-IV)
II Assess Symptom Profile
Beck Anxiety Inventory (cutoff score 10+), Cognitions Checklist—Anxiety subscale (M = 18.13, SD
= 10.06 for primary diagnosis of anxiety disorder),* Penn State Worry Questionnaire (cutoff score
45+), Anxiety Sensitivity Index (M = 19.1, SD = 9.11 for nonclinical; M = 36.4, SD = 10.3 for panic
disorder)+, BDI-II (cutoff score 14+), Daily Mood Rating (Appendix 6.1—Daily Anxiety Ratings and
Situation Record); optional measures (HRSA, DASS, STAI)
III. Immediate Fear Activation Profile
1. Situational Analysis (assess environmental, interoceptive, and cognitive triggers; use Appendix
5.2—Situational Analysis Form; detailed description, rate intensity and duration of anxiety, escape/
avoidance responses, specific triggering cues; begin with in session and then assign as selfmonitoring)
2. Assess First Apprehensive Thoughts (give illustrative explanation on page 142; probe— “What’s
the worst that could happen?”, “What concerns you about the situation?”; use Appendix
5.4—Apprehensive Thought Record to self-monitor; begin with in-session probing)
3. Perceived Autonomic Arousal (typical physiological responses and their interpretation; use Appendix
5.3—Physical Sensation Self-Monitoring Form or Appendix 5.5—Expanded Physical Sensations
Checklist for self-monitoring; in session and self-monitor)
4. Automatic Defensive Responses (probe for automatic cognitive avoidance, reassurance seeking,
compulsions, immediate fight/flight, avoids eye contact, fainting, automatic safety seeking, freezing,
etc.; complete in session and observation)
5. Cognitive Processing Errors (give client list of common errors—Appendix 5.6, and use Identifying
Anxious Thinking Errors to discover client’s typical errors; complete in session)
IV. Secondary Elaborative Response Profile
1. Evaluate Coping Responses (assess behavioral and emotional coping responses when anxious; use
Appendix 5.7—Behavioral Responses to Anxiety Checklist in session)
2. Assess Safety-Seeking Function of Coping Responses (identify responses used to instill sense of
safety and its effects on anxiety; complete in session)
3. Identify Constructive, Adaptive Approaches to Anxiety (any evidence that client has healthy ways of
coping with anxiety in other situations; complete in session)
4. Assess Role of Worry (use Appendix 5.8—Worry Self-Monitoring Form A to assess worry content;
determine its effects on anxiety; complete in session)
5. Identify Cognitive Coping Strategies (use Appendix 5.9—Cognitive Responses to Anxiety Checklist
to identify reliance and perceived effectiveness of maladaptive cognitive responses like thought
suppression, reassurance seeking, thought stopping, etc.; complete in session)
6. Obtain Description of Threat Reappraisal (use Appendix 5.10—Anxious Reappraisal Form to obtain
anxious and nonanxious appraisals; latter becomes goal of treatment; complete in session)
V. Complete Case Formulation (use Appendix 5.11—Diagram of Cognitive Case Conceptualization of
Anxiety)
*Steer, R. A., Beck, A. T., Clark, D. A., & Beck, J. S. (1994). Psychometric properties of the Cognitions Checklist with
psychiatric outpatients and university students. Psychological Assessment, 6, 67–70.
+Antony, M. M. (2001). Measures for panic disorder and agoraphobia. In: M. M. Antony, S. M. Orsillo, & L. Roemer (Eds.),
Practitioner’s guide to empirically based measures of anxiety (pp. 95–125). New York: Kluwer Academic/Plenum.
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
179
Chapter 6
Cognitive Interventions for Anxiety
Courage is not the lack of fear but the ability to face it.
—LT. JOHN B. P UTNAM JR. (23-year-old
American airman killed in World War II)
Pierre is a 33-year-old married man with two preschool children who had a
15-year history of panic disorder and a single episode of major depression in
remission. Past treatment was primarily pharmacotherapy that proved quite
effective in reducing his depression but had less impact on his anxiety symptoms. Pierre was now interested in pursuing a course of CBT for anxiety and
panic symptoms.
At intake Pierre met diagnostic criteria for panic disorder. He reported at
least five full-blown panic attacks in the past month that included heart palpitations, sweating, nausea, shortness of breath, hot flushes, dizziness, and lightheadedness. Nausea was the initial physical sensation that often precipitated a
panic attack. Pierre was fearful that the nausea would lead to vomiting. His
greatest fear was losing control and vomiting in a public setting. As a result he
was hypervigilant for any signs of nausea or abdominal discomfort. He discovered that social situations were more likely to trigger nausea and heightened
levels of anxiety and so he tended to avoid these situations or leave as soon as
he felt abdominal discomfort. Because of his apprehension about heightened
anxiety and panic, Pierre developed limited agoraphobic symptoms in order to
avoid the risk of panic.
The main cognitive basis to Pierre’s anxiety was his belief that “feeling
nausea or abdominal discomfort in a public setting could cause vomiting, or
at least intense anxiety or panic.” His catastrophic misinterpretation of nausea
was not related to a fear of vomiting per se (i.e., he was not fearful of becoming ill), but rather that he would have a panic attack that would cause intense
embarrassment from vomiting in public. He could only recall one incident in
which he vomited in response to a severe panic attack. It appears that this
incident may have been caused by a recent increase in his medication. More
recently there was evidence that the anxiety may be generalizing to other situations such as flying, travel away from home, and sleep.
180
Cognitive Interventions for Anxiety
181
Pierre developed a number of coping strategies to minimize his anxiety.
Although escape and avoidance were his dominant safety-seeking response
style, he carefully monitored what he ate and drank, would sit at the back of a
gathering and close to an aisle, and he always carried his clonazepam with him
whenever he left home. Pierre’s exaggerated appraisal of threat associated with
nausea was not apparent in other areas of his life. He was an avid ice hockey
player who continued to play goalie on a senior men’s team. Thus he regularly
put himself in harm’s way, stopping pucks and often causing significant injury
or pain to himself. This did not make him the least bit anxious. Instead it was
feeling nausea or abdominal discomfort that was associated with appraisals of
unacceptable threat and danger.
Therapy focused on Pierre’s catastrophic misinterpretation of nausea. In
vivo exposure was of limited value because Pierre was already forcing himself
into anxious situations, although he would often leave whenever he became
concerned with nausea. Interoceptive exposure was not utilized because of the
difficulty in producing nausea sensations in a controlled setting. Instead therapy utilized mainly cognitive intervention strategies that targeted Pierre’s faulty
appraisal of nausea, dysfunctional belief that nausea will lead to panic and vomiting, and the belief that escape provided the most effective means of ensuring
safety. Education into the cognitive therapy model of panic, evidence gathering,
generating alternative interpretations, and empirical hypothesis testing were
the primary cognitive intervention strategies employed. After eight sessions,
Pierre reported a significant reduction in panic even with increased exposure
to anxiety-provoking situations. Symptoms of general anxiety showed some
improvement, although to a lesser degree. Therapy continued with a focus on
other issues related to his general level of anxiousness and depressive symptoms
such as low self- confidence and pessimism.
In this chapter we describe cognitive therapy for the maladaptive appraisals and
beliefs that contribute to the persistence of anxiety. We begin with the purpose and
main objectives that underlie cognitive interventions. This is followed by a discussion of
how to educate the client into the cognitive model and teach skills in the identification
of automatic anxious thoughts and appraisals. We then describe the use of cognitive
restructuring to modify exaggerated threat and vulnerability appraisals as well as the
need to eliminate intentional thought control responses. Empirical hypothesis testing is
next described as the most potent cognitive intervention strategy for modifying anxious
cognition. The chapter concludes with a brief consideration of some newer cognitive
interventions such as attentional training, metacognitive intervention, imaginal reprocessing, mindfulness, and cognitive diffusion that appear promising adjuncts in cognitive therapy of anxiety.
MAIN OBJECTIVES OF COGNITIVE INTERVENTIONS
The cognitive treatment strategies outlined in this chapter are based on the cognitive
model of anxiety described in Chapter 2 (see Figure 2.1). They are intended to target the
anxious thoughts, appraisals, and beliefs highlighted in the assessment and case conceptualization (see Chapter 5). Cognitive interventions seek to shift the client’s perspective
from one of exaggerated danger and personal vulnerability to a perspective of minimal
182
ASSESSMENT AND INTERVENTION STRATEGIES
acceptable threat and perceived ability to cope. There are six main objectives of cognitive interventions for anxiety.
Shift Threat Focus
One of the first objectives of cognitive interventions is to shift the client’s focus away
from an internal or external situation or stimulus as the cause of fear and anxiety. Most
individuals with an anxiety disorder enter therapy believing that the cause of their anxiousness is the situation that triggers their anxious episodes. For example, individuals
with panic disorder believe they are anxious because they have chest pain that could
result in a heart attack, whereas individuals with GAD believe the cause of their anxiety
is the real possibility of negative life experiences in the near future. As a result of this
belief, anxious individuals seek interventions that will alleviate what they consider the
source of the anxiety. The person with panic disorder seeks to eliminate chest pain,
thereby removing the possibility of a heart attack, whereas the person with social phobia may look for signs that he is not being negatively evaluated. One of the first tasks in
cognitive therapy is to guide clients into an acknowledgment that the situational triggers
and perceived possibilities of terrible outcomes is not the cause of their anxiety. This
is accomplished through the cognitive restructuring and empirical hypothesis-testing
interventions that are discussed below.
It is critical that the cognitive therapist avoid any attempt to verbally persuade anxious clients against their anxious threat. This warning against trying to verbally modify
threat content was emphasized by Salkovskis (1985, 1989) for treatment of obsessions.
Thus the therapist must not engage in verbal debates about the possibility of having a
heart attack, suffocating, contaminating others with a deadly germ, making a mistake,
being negatively evaluated in a social setting, being the victim of another assault, or
experiencing some negative outcome in the future. After all, any clever arguments that
can be concocted by the therapist will be immediately dismissed by the client because
mistakes do happen, people can become the victim of disease by contamination, and
even the occasional young person dies from a heart attack. The reality is that threat can
never be eliminated entirely. At best such persuasive debates will only amount to reassurance that provides temporary relief from anxiety and at worst the client’s outright
dismissal of the effectiveness of cognitive therapy. Thus it is critical to the success of
cognitive therapy that therapy avoids a direct focus on the client’s threat content.
Clinician Guideline 6.1
Avoid any attempt to use logical persuasion to directly target primary threat content. Such
attempts will undermine the effectiveness of cognitive therapy and result in the persistence
of the anxious state.
Focus on Appraisals and Beliefs
The cognitive perspective views anxiety in terms of an information-processing system
that exaggerates the probability and severity of threat, minimizes personal ability to
cope, and fails to recognize aspects of safety (i.e., Rachman, 2006). An important objec-
Cognitive Interventions for Anxiety
183
tive in cognitive therapy, then, is to shift the client’s focus from threat content to how
they appraise or evaluate threat. For cognitive therapy to be effective, the client must
accept the cognitive model (i.e., treatment rationale) that their anxiety arises from their
faulty thoughts, beliefs, and appraisals of threat rather than from the threat content
itself.
This approach to anxiety recognizes that individuals with an anxiety disorder often
fail to adopt a rational, realistic appraisal of the dangers related to their anxious concerns, especially during anxious states. In fact anxious individuals often recognize that
a danger is highly unlikely, or even impossible. However, the problem is that they will
appraise even a remote danger (1/1,000,000,000) as an unacceptable risk. Thus the
cognitive therapist must focus on the thoughts, appraisals, and beliefs about threat (e.g.,
feelings of nausea) and vulnerability rather than threat content per se. The following is a
clinical vignette that illustrates how this shift in therapeutic orientation can be achieved
with a person suffering from social phobia:
THERAPIST: Looking over your diary, I see that you were especially anxious in a meeting
you had with work colleagues last week.
CLIENT: Yes, the anxiety was really intense. I was so scared someone would ask me a
question.
THERAPIST: What would be so bad about that?
CLIENT: I’m afraid I would say something stupid and everyone would think I’m an
idiot.
THERAPIST: What do you think was making you anxious about the meeting?
CLIENT: Well I was anxious because I could be asked a question and then I would say
something stupid and everyone would think badly of me. [focus on threat content]
THERAPIST: It sounds like you certainly had anxious thoughts like “what if I’m asked a
question” and “what if I say something stupid.” Do you suppose other people who
do not have social anxiety also have these same thoughts from time to time?
CLIENT: Well, I suppose they do but I feel so anxious and they don’t.
THERAPIST: True, that is an important difference. But I wonder if this difference is
caused by how you evaluate these thoughts when you have them and how a nonanxious person evaluates the thoughts when she has them about a work meeting.
CLIENT: I’m not sure I understand what you mean.
THERAPIST: When you think “I could be asked a question” and “I could say something
stupid,” how likely do you think this is and what do you think could be a consequence or outcome?
CLIENT: When I’m anxious I tend to be entirely convinced I’m going to say something
stupid and that everyone will think I’m an idiot.
THERAPIST: So when you have these anxious thoughts you evaluate the probability that
it will happen as very high (“you will say something stupid”) and that terrible consequences will result (“everyone will think I’m an idiot”). Do you suppose this might
be the source of your anxiety, that it is these appraisals of high probability and serious consequences that are making you anxious? [focus on appraisals of threat]
184
ASSESSMENT AND INTERVENTION STRATEGIES
CLIENT: Well, I don’t really know. I always thought that what made me anxious is that
I tend to say stupid things when I’m around people.
THERAPIST: Let’s see if we can find out more about this. For a homework assignment,
do you have some close friends or family you could ask about whether they have
ever had concerns about saying something stupid in a public setting? It would be
interesting to find out how they appraise or think about these situations that results
in not feeling anxious.
CLIENT: Yes, I could do that.
THERAPIST: Great! So let’s see whether the way we appraise or think about situations
(e.g., “I will probably say something stupid and everyone will think I’m an idiot”)
is an important cause of anxiety or not. If these appraisals are important, then we
will want to change them as part of our treatment for social anxiety.
Clinician Guideline 6.2
A key element of cognitive therapy of anxiety is teaching clients that the source of persistent anxiety is their biased appraisals of threat. The success of other cognitive interventions
depends on clients’ acceptance of this cognitive or information-processing formulation of
anxiety.
Modify Biased Threat, Vulnerability, and Safety Appraisals and Beliefs
In cognitive therapy of anxiety the main objective of cognitive interventions is to modify
overestimated appraisals of threat and personal vulnerability related to the primary
anxious concern as well as change the client’s perspective on the safety aspects of the
situation. Cognitive interventions tend to focus on four key elements of faulty cognition.
•฀ Probability estimates: What is the perceived threat or danger? Is the client generating an exaggerated probability estimate of the threat or danger?
•฀ Severity estimates: Is there a biased evaluation of the severity of the perceived
outcome or consequence of the threat?
•฀ Vulnerability estimates: What is the level of perceived personal vulnerability
when in the anxious situation? To what extent are the client’s perceived weaknesses exaggerated when anxious?
•฀ Safety estimates: What safety information is being ignored or undervalued, resulting in a downgraded estimate of perceived safety in the anxious situation?
The faulty appraisals of threat and vulnerability are evident in the automatic apprehensive thoughts or images, misinterpretations of physiological arousal, cognitive errors,
dysfunctional defenses and coping strategies, and primary worry symptoms identified in
the case conceptualization (see Appendix 5.11). Table 6.1 illustrates typical appraisals
that are associated with the anxiety disorders.
Once the biased appraisals have been well articulated in therapy, the goal of cognitive interventions is to arrive at a more balanced, realistic appraisal of the probability
Cognitive Interventions for Anxiety
185
TABLE 6.1 Illustrative Examples of Threat, Vulnerability, and Safety Appraisals Associated
with the Anxiety Disorders
Perceived
vulnerability
estimates
Anxiety
disorders
Threat probability
appraisals
Threat severity
appraisals
Panic
disorder
“I’m having difficulty
breathing; I’m not
getting enough air.”
“What if I can’t
breathe and suffocate
to death?”
“I can’t handle this
feeling of not being
able to breathe;
it is a terrifying
experience.”
“No one is around
to help me. I’m so
far from a hospital. I
need more oxygen.”
Generalized
anxiety
disorder
“I just know that I’m
going to do poorly in
the job interview.”
“I’ll make such a
fool of myself; the
interviewers will
wonder why I ever
applied for this job.
I’ll never find a good
job.”
“I never interview
well. I become so
anxious that I lose
my concentration and
end up rambling all
over the place.”
“Job interviewers are
just looking for an
excuse to reject you.
Besides they have
already made up their
mind not to hire you
before you start the
interview.”
Social phobia
“People are looking
at me and notice that
I’m shaky.”
“They’ll wonder
what’s wrong with
me; does she have a
mental illness?”
“I can’t cope
with these social
situations; I get too
anxious.”
“I can’t conceal my
anxiety from others;
how could anyone not
see that I’m anxious.”
Obsessive–
compulsive
disorder
“I have a terrible
feeling that I didn’t
turn off the stove.”
“If I did leave the
stove burner on, it
could start a fire.”
“I am prone to
making mistakes,
being forgetful, and
so could easily leave
the burner on.”
“I don’t have an
accurate memory of
turning it completely
off. I need to check
and concentrate hard
on whether the knob
is completely off.”
Posttraumatic
stress
disorder
“I have to avoid
situations that remind
me of the trauma
because I will have
intrusive recollections
of what happened to
me.”
“I feel so helpless,
alone, and frightened
when I have these
intrusive thoughts
and memories of the
ambush. It’s almost
as bad as when I was
under fire.”
“I’ve got to stop
having these intrusive
thoughts and
flashbacks of the
ambush. And yet I
can’t control them;
they’ve taken over my
life.”
“The only time I can
forget is when I’m
drinking. There is
no escape from the
memories even when
I’m asleep.”
Biased safety
estimates
and severity of threat, the person’s actual ability to cope with the situation, and whether
it is more realistic to assume safety rather than danger. This latter perspective can only
be achieved by helping clients abandon their maladaptive safety-seeking practices and
focus on aspects of the anxious situation that denote safety. Interventions such as cognitive restructuring and empirical hypothesis testing are used to achieve this modification
in anxious thoughts, beliefs, and appraisals.
A focus on the modification of threat appraisals has always been at the heart of
cognitive therapy for anxiety (e.g., D. M. Clark, 1986b; Wells, 1997). Beck et al. (1985,
2005) state that cognitive restructuring teaches clients to replace questions about “why”
they are feeling anxious with “how” they are making themselves feel anxious (i.e.,
appraisals of threat). Recent cognitive behavioral treatment manuals for the anxiety disorders have also emphasized the use of cognitive interventions to modify threat apprais-
186
ASSESSMENT AND INTERVENTION STRATEGIES
als (e.g., Craske & Barlow, 2006; D. A. Clark, 2004; D. M. Clark, 1997; Rachman,
2003; Rygh & Sanderson, 2004; Taylor, 2006). In addition, evidence from the social
experimental literature on emotion regulation indicates that cognitive reappraisal as a
coping strategy is associated with greater positive emotion, less negative emotion, and
better psychological health (John & Gross, 2004). Thus our emphasis on the reappraisal
of threat and vulnerability has broad support in the psychotherapeutic and experimental literature.
Clinical Guideline 6.3
The primary focus of cognitive interventions is the modification of exaggerated estimates
(appraisals) of the probability and severity of threat as well as evaluations of personal vulnerability and lack of safety.
Normalize Fear and Anxiety
Normalizing anxiety was first discussed by Beck et al. (1985) in their chapter on modifying the affective component of anxiety. At that time normalizing anxiety was highlighted as a way to help clients become less self-absorbed in their anxiety symptoms.
There are three aspects of the normalization of anxiety that must be considered.
1. Normalizing in relation to others. The actual situations, thoughts, and sensations that are associated with anxiety should be normalized. Anxious individuals are
often so focused on their own experience of anxiety that they fail to recognize that these
phenomena are almost universal. For example, how often do people experience chest
pain or breathlessness, a concern that they have made a bad impression on others, doubt
over their actions or decisions, uncertainty about the possibility of some accident or
future calamity, or recollections about some frightening experience? The therapist can
ask clients to consider the “normality of threat” and possibly even collect survey data
on whether nonanxious individuals ever experience the anxious threat. The purpose of
this exercise is to shift individuals’ focus away from threat content as the source of their
anxiety to their appraisal of threat as the main contributor to their anxious state.
2. Normalizing in relation to past experiences. The therapist should explore clients’ past experience with the situations, thoughts, or sensations that now trigger their
anxiety. “Was there a time when having tightness in your chest didn’t really bother
you?” “Have you always been so concerned about what others think of you?” “Was
there a time when concern about germs was not such a big deal in your life?” By inquiring about their past, clients will be remembering a time when they coped much better
with the perceived threat. Again this shifts the focus from “I am an anxious person” to
“What am I doing now that has made my anxiety so much worse?”
3. Normalizing in relation to situations. When assessing the situations that trigger
anxiety, the cognitive therapist can also identify other situations that trigger the same
thoughts or sensations but that do not lead to an anxious episode. For example, when
working with panic disorder it is often helpful to inquire whether the client experiences
physical sensations when exercising or engaging in vigorous activity but does not feel
Cognitive Interventions for Anxiety
187
anxious. In fact clients could be asked to exercise as a behavioral experiment to highlight
their different appraisals of physical sensations (see discussion in next chapter). This
type of normalization highlights the situational nature of anxiety and again emphasizes
the client’s ability to cope with anxiety-related triggers when they occur in nonanxious
situations. It also reinforces the cognitive perspective that anxiety arises from appraisals
rather than the actual stimuli that trigger anxiousness. (E.g., “When you are exercising
and you feel tightness in your chest, you attribute this to physical exertion. You expect
to feel tense while exercising. But when you feel spontaneous chest tightness, you attribute this to a possible impending heart attack. You tell yourself something is wrong,
this shouldn’t be happening. So when exercising you interpret chest tightness in a way
that results in no anxiety, whereas when the chest tightness arises unexpectedly, you
interpret the sensations in another way that leads to anxiousness, even panic.”)
Normalizing fear and anxiety is an important objective in cognitive therapy of
anxiety. It not only reinforces the focus on threat appraisals as the source of anxiety, but
it produces a more optimistic attitude toward overcoming anxiety. Clients are reminded
that very often they react to threat in a nonanxious, even courageous manner. As Rachman (2006) recently noted, “In specifiable circumstances virtually everyone, including
patients suffering from anxiety disorders, can behave courageously” (p. 7). In cognitive
therapy we remind clients that they often “turn off the fear program” in a variety of
situations not related to their anxiety disorder. The goal of treatment, then, is to build
on their own natural abilities to overcome fear and apply these resources to the anxiety
disorder.
Clinician Guideline 6.4
Normalization of fear and anxiety, an important element of cognitive therapy, is achieved by
emphasizing the universality of threat, the client’s past experiences with anxious cues, and
the situational or variable nature of anxious triggers.
Strengthen Personal Efficacy
In cognitive therapy therapeutic interventions do not focus only on modifying faulty
threat appraisals but also on correcting erroneous beliefs about personal vulnerability
and perceived inability to deal with one’s anxious concerns. The cognitive therapist
can construct the client’s vulnerability perspective from the first apprehensive thoughts,
automatic defensive responses, coping strategies, and worries identified in the cognitive
case conceptualization. An important theme that runs throughout the course of treatment is “You’re stronger than you think” when it comes to dealing with the anxious
concerns. Building a greater sense of self-efficacy (i.e., Bandura, 1977, 1989) by structuring experiences and highlighting information that reinforces perceived control or
mastery of the anxiety-related threat are critical elements in cognitive therapy of anxiety
that will help clients override the threat-schema activation.
During cognitive restructuring and empirical hypothesis-testing exercises the cognitive therapist emphasizes the difference between an initial vulnerability estimate and
188
ASSESSMENT AND INTERVENTION STRATEGIES
the actual outcome related to an anxious situation. The goal is to teach clients how
their initial thoughts and beliefs about vulnerability are a faulty representation of reality that makes them more anxious and contributes to avoidance and ineffective coping
responses. The following clinical vignette illustrates how perceived vulnerability can be
challenged with a client suffering from generalized anxiety.
CLIENT: I have been worried for a few days about my daughter’s visit. I am so concerned
that everything will go well. You know I haven’t seen her for so long. When she left
home a couple of years ago we had such a big argument. At that time she swore she
would never come back home again.
THERAPIST: What’s the worst that could happen when she visits?
CLIENT: Well, she could bring up the past and then we would get into a huge argument.
She would then storm out of the house and never return.
THERAPIST: That would certainly be a terrible outcome for you. I know how much you
really love your daughter.
CLIENT: Yeah, I’ve been trying to think how I can avoid an argument.
THERAPIST: And what have you come up with?
CLIENT: Basically nothing. Every time I try to visualize how it will go and what I will do
if she brings up the past, all I can see is anger, shouting, and her slamming the door
as she leaves the house. [low self- efficacy appraisals and beliefs]
THERAPIST: Sounds like you feel pretty helpless. When you are thinking like this what
happens to your anxiety and worry?
CLIENT: I just end up feeling more anxious and worried about the visit.
THERAPIST: So one effect of thinking that you are incapable of handling this situation is
that your anxiety and worry escalate. How do you think all of this will affect your
interactions with your daughter?
CLIENT: I don’t think it is helping me in any positive way. I end up feeling so scared and
confused, probably I will end up blurting out something stupid when she is with me
that will only make matters worse.
THERAPIST: Okay, let me summarize. You’ve described worries over your daughter’s
visit next weekend. One of the themes running through this worry is “I’m helpless
to avoid a conflict” and this helplessness makes you feel even more anxious and less
prepared for your daughter’s visit. But I wonder if you are as helpless as you think.
I wonder if you as poor at coping with confrontation or your daughter’s anger as
you think. I would like to suggest a couple of things. First, let’s go over some of
your past experiences with people who are angry or confrontational and see how
you’ve managed. Are you as bad at dealing with these situations as you think?
And second, let’s take a problem-solving approach and write down, maybe even
role-play, some strategies you might use with your daughter when she visits. [The
therapeutic intervention seeks to contrast the client’s predicted self- efficacy with
actual outcomes in the past in order to highlight discrepancy and exaggeration of
low perceived self-efficacy.]
CLIENT: This sounds like a good idea. I’m really worried about this visit.
Cognitive Interventions for Anxiety
189
Clinician Guideline 6.5
The therapist focuses on correcting low perceived self-efficacy for anxiety by pointing out
how a discrepancy between predicted ability to cope and actual past outcomes contributes
to anxiety. In addition the therapist adopts a problem-solving approach to expand the client’s repertoire of adaptive coping resources and to foster positive experiences to enhance
self-efficacy.
Adaptive Approach to Safety
In Chapter 3 we reviewed empirical research indicating that safety-seeking thoughts,
beliefs, and behaviors are important contributors to anxiety. Consequently dealing with
safety-seeking issues is an important theme in CT for anxiety. Three aspects of safety
seeking should be considered in treatment.
Faulty Risk Appraisals
Salkovskis (1996a) noted that threat appraisal that leads to safety seeking is a balance between the perceived probability and severity of threat, on the one hand, and
coping ability and perceived rescue factors, on the other. Kozak, Foa, and McCarthy
(1988) commented that in OCD danger is assumed unless there is evidence for complete
safety whereas the opposite viewpoint prevails in nonanxious states in which safety is
assumed unless there is valid evidence of danger. The person with panic disorder may
find heart rate increases too risky, or the person with OCD might be convinced that any
observable dirt is a harbinger of disease and destruction. This strategy will confirm the
patient’s fear while disconfirming safety evidence is overlooked.
An important goal of cognitive therapy is to investigate with clients whether they
hold faulty appraisals and assumptions about risk. What, then, constitutes “an acceptable level of risk”? “Can one eliminate all possibility of risk?” “What effect does this
have on a person’s life?” “Do nonanxious people live with risk”? “How successful have
you been at eliminating all risk and at what cost to you?” These are questions that the
cognitive therapist explores with clients when reviewing their self-monitoring diaries in
an effort to correct maladaptive risk appraisal.
Enhance Safety-Seeking Processing
There are many aspects of anxious situations that signal safety rather than threat, but
the anxious person often misses this information. When reviewing homework assignments, attention can be drawn to safety elements that the client may have ignored or
minimized. Furthermore, anxious clients can be asked to intentionally record any safety
information conveyed in an anxious situation. This safety information can be contrasted
with threat information in order to generate a more realistic reappraisal of the magnitude of the risk associated with a particular situation. Throughout treatment the cognitive therapist must be vigilant for biases that minimize safety and maximize threat,
thereby resulting in a threat-oriented information processing bias.
190
ASSESSMENT AND INTERVENTION STRATEGIES
Dysfunctional Avoidance and Safety-Seeking Behavior
An important objective in cognitive therapy for anxiety is the identification and subsequent correction of avoidance and maladaptive safety-seeking behavior that contributes
to the persistence of anxiety. As noted in the cognitive case conceptualization, these
safety-seeking strategies can be cognitive or behavioral in nature. For example, clients
with panic disorder might use controlled breathing whenever feeling breathless in order
to avert a panic attack, or the person with social anxiety may avoid eye contact in social
interactions.
Often safety-seeking responses have been built up over many years and may occur
quite automatically. In such cases one can not expect the client to immediately cease
the safety-seeking behavior. Instead the cognitive therapist should challenge the safetyseeking gradually, first working with the client to understand the role of such behavior
in the persistence of anxiety. Once the client acknowledges its deleterious effects, then
the maladaptive coping can be gradually phased out and substituted with more positive
adaptive strategies. It is likely that this process may have to be repeated a number of
times for anxious clients with multiple avoidant and safety-seeking responses.
Clinician Guideline 6.6
The clinician must address faulty risk appraisals, inhibited processing of safety cues, and
maladaptive avoidant and safety-seeking responses throughout the course of cognitive
therapy of anxiety disorders. Gradually phase out maladaptive safety-seeking responses and
replace them with alternative, more adaptive strategies over an extended period of time.
COGNITIVE INTERVENTION STRATEGIES
In this section we present the actual therapeutic strategies that can be used to achieve
the main objectives of cognitive therapy for anxiety. Naturally, these intervention strategies will be modified when used with the specific anxiety disorders discussed in the third
part of this volume.
Educating the Client
Educating clients has always played a central role in cognitive therapy (Beck et al., 1979,
1985, 2005). Today it continues to be emphasized in practically every cognitive therapy
and cognitive-behavioral treatment manual (e.g., J. S. Beck, 1995; D. A. Clark, 2004; D.
M. Clark, 1997; Craske & Barlow, 2006; Rygh & Sanderson, 2004; Rachman, 1998,
2003, 2006; Taylor, 2006; Wells, 1997). The didactic component of treatment may not
only improve treatment compliance but it can also directly contribute to the correction
of faulty beliefs about fear and anxiety (Rachman, 2006).
There are three aspects of educating the client that are important in cognitive therapy for anxiety. First, individuals often have misconceptions about anxiety and so a
discussion of fear and anxiety should be given with reference to the client’s personal
experiences. Second, a cognitive explanation for the persistence of anxiety should be
Cognitive Interventions for Anxiety
191
TABLE 6.2. Primary Elements of Educating the Client into the Cognitive
Model and Treatment of Anxiety
Themes emphasized when educating the client
•฀ Define anxiety and the role of fear
•฀ The universal and adaptive nature of fear
•฀ Cognitive explanation for inappropriate activation of the anxiety program
•฀ Consequences of inappropriate activation of anxiety
•฀ Escape, avoidance, and other attempts to control anxiety
•฀ Treatment goal: turning off the anxiety program
•฀ Treatment strategies used to deactivate the anxiety program
•฀ The role of other approaches to anxiety reduction (e.g., medication,
relaxation, herbal remedies)
provided in a manner that clients can readily understand and apply to their own situation. And third, the cognitive treatment rationale should be clarified so that clients
will fully collaborate in the treatment process. In our experience clients who terminate
therapy within the first three to four sessions often do so because they have not been
educated into the cognitive model or they fail to accept this explanation for their anxiety. Either way, educating the client begins at the first session and will be an important
therapeutic ingredient in the early sessions.
Table 6.2 presents the main themes that should be addressed when educating the
client about the cognitive approach to anxiety. We briefly discuss how the therapist can
communicate this information to clients in a comprehensible manner.
Defining Anxiety and Fear
Clients should be provided with an operational definition of what is meant by fear and
anxiety from a cognitive perspective. Based on the definitions in Chapter 1, fear can be
described as perceived threat or danger to our safety or security. Clients can be asked for
examples of when they felt fearful and what the perceived danger was that characterized
the fear (e.g., near accident, waiting for results of medical tests, threatened with violence
or aggression). It should be pointed out that even thinking about or imaging worst-case
scenarios can elicit fear. Again examples of imagined fears could be discussed. In the
same way anxiety can be described as a more complex, prolonged feeling of unease or
apprehension involving feelings, thoughts, and behavior that occurs when our vital
interests are threatened. Whereas fear is usually momentary, anxiety can last for hours,
maybe even days. Given the ubiquitous nature of computers and information technology
in modern society, most people will readily understand if anxiety is described as analogous to “a computer program that gets turned on, takes over the operating system,
and won’t quit until it is deactivated or turned off.” Throughout treatment, we find it
useful to refer to “activating and deactivating the fear program” and the importance of
“turning off the fear program” in order to eliminate anxiety. The therapist should be
asking the client for personal examples of fear and anxiety in order to reinforce a full
understanding of the concepts. This will ensure that client and therapist have a common
language when talking about experiences of anxiety.
192
ASSESSMENT AND INTERVENTION STRATEGIES
Adaptive Value of Fear
Most individuals suffering from an anxiety disorder have forgotten about the important
role that fear plays in our survival. The therapist should discuss the universal nature
of fear and its survival function. Clients can be asked about times when being afraid
“saved their life” by mobilizing them to deal with a potential threat or danger. Beck et
al. (1985, 2005) noted that it is often helpful to discuss with clients the “fight-or-flight”
response that characterizes fear.
In the same way mild to moderate levels of subjective anxiety (nervousness) can be
adaptive if it is not too intense or prolonged. Being nervous about an impending exam
or job interview might motivate a person to be better prepared. Performers acknowledge
that some degree of nervousness is both expected and beneficial before going on stage.
Again the therapist can solicit past experiences from the client when anxiety was actually functional.
The reason for including a discussion on the positive function and adaptive value
of fear and anxiety is to emphasize that these states are not abnormal. The problem
in anxiety disorders is not the experience of fear or anxiety, but the fact that the fear
program is inappropriately activated or turned on. Thus the goal of therapy is not to
eliminate all anxiety but rather to reduce anxiety that is inappropriate or maladaptive.
Another reason for emphasizing the survival value of fear is to normalize clients’ anxiety so they view it as an exaggeration or misapplication of normal emotion. This should
bolster a great sense of hope and optimism in treatment since they are not as different
from “normal people” as they may have been thinking.
Cognitive Explanation for Inappropriate Activation of Anxiety
The preceding discussion on the normality of fear and anxiety will naturally lead into
the issue of why the client’s anxiety is so much more intense, persistent, and triggered by
things that don’t bother most people. This is the crux of the educational phase because
it is critically important to the success of therapy that clients realize that their appraisals of threat are the primary determinants of their clinical anxiety. A copy of Figure
6.1 can be given to clients in order to facilitate an explanation of the cognitive model of
anxiety.
Education into the cognitive model will occur after the assessment so the therapist
can draw on the cognitive case conceptualization to obtain examples of the client’s typical responses when anxious. The therapist should go through each step in Figure 6.1
and elicit from the client examples of typical situations, automatic thoughts, anxious
symptoms, search for safety and avoidance, worry and preoccupation with anxiety and
helplessness, and failed attempts to control anxiety. These experiences could be written
down on Figure 6.1 as a record for the client on how the cognitive model explains inappropriate fear activation and the persistence of her clinical anxiety. Any questions or
doubts concerning the applicability of the cognitive explanation for the client’s anxiety
should be addressed using the guided discovery in which the therapist questions the client in a manner that will encourage her to reevaluate her misgivings about the cognitive
explanation (Beck et al., 1979). In most cases it is helpful to assign a homework assignment such as having the person fill in Figure 6.1 immediately after an anxiety episode.
This will help consolidate a better understanding and acceptance of a cognitive explanation for the clinical anxiety state.
Cognitive Interventions for Anxiety
Experience certain
situations,
information, or cues
Heightened
attention to
threat and automatic
apprehensive
thoughts
[appraisals of threat]
193
Search for safety
and reduction in
anxiety
ANXIETY
Attempts to
control
anxiety
SYMPTOMS
ACTIVATED
FEAR
Preoccupation
with anxiety and
helplessness;
threat reevaluated
FIGURE 6.1. Diagram of the cognitive model of anxiety for use with clients.
Consequences of Inappropriate Anxiety
Most individuals with an anxiety disorder are all too familiar with the negative consequences of their anxiety. However, it is important to discuss consequences because having “fear of anxiety” is a prominent feature of clinical anxiety (Beck et al., 1985, 2005;
D. M. Clark, 1986b). The therapist can explore with the client whether being “anxious
about being anxious” might actually intensify the clinical disorder by making a person
more sensitive or vigilant for any signs of anxiety (i.e., latter phase in Figure 6.1).
It is important to discuss how anxiety is manifested in the three major response
systems; the physiological, the behavioral, and the cognitive. This should be discussed
in reference to the clients’ own experience of anxiety. Craske and Barlow (2006) provide
a very helpful explanation of the three components of anxiety in their self-help book for
worry called Mastery of Your Anxiety and Worry. They note that a better understanding of the physical, cognitive, and behavioral components of anxiety helps reduce the
mystery and uncontrollability of anxiety and provides a framework for learning ways
to reduce anxiety.
Some discussion of the broader consequences of having anxiety should be incorporated into educating the client. What effect does anxiety have in the client’s daily life
at work, home, and leisure? Are there restrictions or limitations imposed on what individuals can do or where they can go? The broader negative impact of anxiety needs to
be emphasized in order to encourage client commitment to the therapeutic process by
helping individuals think in terms of the costs and benefits of change. A consideration
of the “personal burden of anxiety” can also help in the establishment of treatment
goals.
The Role of Avoidance and Safety Seeking
It is useful to ask clients what they think is the most effective way to reduce anxiety.
Although a variety of answers may be given, it should be emphasized that escape and
avoidance (or completion of a compulsive ritual in OCD) ensure the quickest reduction
194
ASSESSMENT AND INTERVENTION STRATEGIES
in anxiety. The therapist and client can discuss a number of life-threatening examples
where escape or avoidance actually ensures one’s survival. Examples can also be given
of animals (i.e., the client’s pets) that automatically escape or avoid perceived danger.
It should be emphasized that escape and avoidance are natural responses to perceived
threat and danger.
A discussion of the natural, automatic character of escape and avoidance should
lead into a consideration of their negative consequences and how escape and avoidance contribute to the persistence of anxiety. In their self-help book on panic entitled
10 Simple Solutions to Panic, Antony and McCabe (2004) cite four disadvantages of
escape/avoidance:
•฀ It prevents learning that situations are safe, not dangerous or threatening (i.e.,
failure to disconfirm faulty appraisals and beliefs of threat).
•฀ The subjective relief associated with escape/avoidance reinforces this behavior in
future episodes of anxiety.
•฀ Giving into escape/avoidance will increase a sense of guilt and disappointment in
one’s self and a loss of self- confidence.
•฀ The immediate relief associated with escape/avoidance increases one’s sensitivity
to threat cues so that in the long term it will maintain or even increase fear and
anxiety.
Throughout this discussion of the negative effects of escape/avoidance, the therapist
should be soliciting personal examples and questioning the client on any perceived
adverse consequences of continued escape/avoidance. By educating the client on the role
of escape/avoidance in anxiety the therapist seeks to increase awareness that elimination of this control strategy is critical to the success of treatment. It will also lay the
groundwork for introducing prolonged exposure to threat as the obvious remedy for
this maladaptive defensive strategy (a fact that most individuals with anxiety are most
reluctant to accept).
The therapist should also explore with clients any dysfunctional safety-seeking
behaviors that may be used to alleviate anxious feelings. Do clients carry anxiolytic
medications at all times just in case they are needed? Do they only venture into certain
places when accompanied by a close friend or family member? Are there other more
subtle forms of safety seeking such as holding onto railings when feeling dizzy or automatically sitting down when feeling weak? After examples of safety seeking are elicited,
the therapist should discuss how this form of coping with anxiety might contribute to
its persistence because:
•฀ It prevents one from learning that his fears (i.e., perceived threats) are groundless
(Salkovskis, 1996a).
•฀ It creates a false sense of security (e.g., person with panic disorder develops maladaptive belief that having a friend close by somehow reduces the risk of heart
palpitations and a heart attack).
Once again the purpose for educating clients about the role of safety-seeking responses
is to increase their acceptance that reduction in this behavior is an important goal of
treatment.
Cognitive Interventions for Anxiety
195
Treatment Goal
In keeping with our metaphor of fear as “a computer program,” the therapist introduces
the treatment rationale by explaining the goal of cognitive therapy in terms of “deactivating or turning off” the fear program by deliberately and intentionally engaging in
activities that will “override” or “counter” fear and anxiety. The therapist should refer
to Figure 6.1 and indicate that the fear program can be deactivated by intervening at all
the different steps that contribute to the persistence of anxious symptoms. Clients could
be asked to provide examples of their own success in deliberately overcoming an initial
fear. It is also important to question the client about treatment expectations in order to
elicit any misconceptions that could undermine the success of cognitive therapy.
There are a number of common faulty misconceptions about treatment that might
need addressing. First, treatment can not permanently shut down fear. The goal is not
to eliminate anxiety totally (if that was even possible) but to help clients develop effective ways to override the fear program when it is inappropriately activated. Second, the
experience of anxiety will feel more natural, whereas efforts to reduce anxiety will seem
much more difficult. This is because the former is an automatic response to perceived
threat and the latter requires a much more deliberate, effortful response. This does not
mean that intentional responses to anxiety are not powerful enough to deactivate fear
and reduce anxiety. What it does mean is that repeated experiences with these effortful responses will be needed in order to improve their efficiency and effectiveness. And
third, the objective of cognitive therapy is not to teach people more effective ways to
“control their anxiety.” Instead cognitive therapy focuses on helping individuals develop
a more “accepting attitude” toward anxiety rather than a “combative (i.e., controlling)
attitude.” When thoughts like “I can’t let these anxious feelings continue” are replaced
with “I can allow myself to feel anxious because I know I’m exaggerating the threat and
danger,” then the intensity and persistence of anxiety are greatly diminished (Beck et
al., 1985, 2005).
Treatment Strategies
Clients should be provided with a brief description and rationale for the intervention
strategies that will be used to “turn off” the fear program and diminish their anxious
feelings. The therapist should explain that a greater understanding of one’s anxiety
through education and the self-monitoring of anxious episodes are important interventions in cognitive therapy of anxiety. These components of treatment help counter the
unexpected and unpredictable nature of anxiety.
The therapist explains that a second class of cognitive therapy interventions focuses
directly on changing anxious thoughts and beliefs. This is accomplished by learning to
critically question whether one’s initial apprehensive thoughts are an accurate appraisal
of the situation and then replacing these anxious interpretations with a more realistic
way of thinking. Specific behavioral experiments are designed that will help the client
develop a less anxious way of thinking. The therapist should emphasize that developing
new ways of thinking about their anxious concerns is an important part of treatment
because it directly targets the automatic apprehensive thoughts that give rise to anxious
symptoms (refer to Figure 6.1).
A third category of cognitive therapy interventions deals with behavioral responses
and coping strategies that may contribute to the persistence of anxiety. Escape, avoid-
196
ASSESSMENT AND INTERVENTION STRATEGIES
ance, safety-seeking behavior, and other cognitive or behavioral responses employed by
clients in an effort to control their anxiety are targeted for change. Alternative ways of
responding to anxiety are introduced and clients are encouraged to evaluate the utility
of these approaches through use of behavioral exercises.
A final ingredient of cognitive therapy for anxiety involves graduated and repeated
exposure to anxiety-provoking situations and a phasing out of escape, avoidance, safety
seeking, or other forms of neutralizing responses (e.g., compulsive rituals in OCD).
When introducing the concept of fear exposure, it must be realized that this can be terrifying to anxious individuals. Many anxious clients refuse to continue with treatment
at the mere mention of exposure because they can not imagine dealing with the intense
anxiety they expect to experience in highly fearful situations. To counter the client’s
negative expectations, the therapist should emphasize that exposure to fear situations
is the most potent intervention for achieving lasting fear reduction. Exposure exercises
will be introduced later in therapy in a very gradual fashion starting with experiences
with a low to moderate level of anxiety in order to elicit core cognitions that underlie
anxious feelings. All assignments will be discussed in a collaborative fashion with the
client having the final say on what is expected at any point in therapy. The therapist
should also reassure clients that an exposure task that seems too difficult can always
be broken down or modified to reduce the level of anxiety. Finally, the therapist should
explain the benefits of exposure to anxious situations. It reduces anxiety by providing
evidence against threat-related “hot” cognitions and beliefs, it bolsters self- confidence,
and it provides opportunity to practice more adaptive ways of coping with anxiety.
Other Approaches to Anxiety
Often clients will inquire whether medication, meditation, herbal remedies, and the
like can be used while having a course of cognitive therapy for anxiety. However, these
approaches are somewhat counterproductive to cognitive therapy because they all emphasize the short-term reduction and avoidance of anxious symptoms without concomitant change in cognition. For many individuals these interventions may have become an
important part of their coping strategy for anxiety. Thus any withdrawal of these interventions should be done gradually, commensurate with a reduction in the client’s anxiety
level with progress through cognitive therapy. Naturally no change in medication should
be recommended unless prescribed by the client’s medical practitioner.
Methods of Educating the Client
Although a certain amount of verbal teaching is an evitable part of the educational process, it should not be the sole means of communicating the cognitive model and treatment rationale. The therapist should be asking clients about their personal experiences
and using guided discovery to emphasize key aspects of the cognitive model that can be
identified in these experiences. Clients are much more likely to accept the model if it has
immediate relevance to their own experiences with anxiety.
The therapist can also assign self-monitoring homework to encourage the client to
explore whether different aspects of the cognitive model are relevant to his anxiety. For
example, a client with social phobia could be asked to experiment with the effects of giving eye contact versus avoiding eye contact in social interactions as a way of determin-
Cognitive Interventions for Anxiety
197
ing whether subtle forms of avoidance and safety seeking have an effect on her anxiety
level. A client with OCD could be asked to try hard to suppress an anxious obsession on
one day and then relinquish control efforts on an alternative day and record the effects
of trying to control anxiety. A person with panic disorder could be asked to record the
effects of thinking about a heart attack when his chest feels tight versus thinking that it
is muscle strain. Notice that all of these assignments focus more on highlighting some
aspect of the cognitive model in the client’s experience of anxiety rather than directly
modifying thoughts or behavior.
Bibliotherapy is an important method of educating the client into the cognitive
model. We are currently in the process of writing a client workbook based on the present volume that will provide explanations and case examples useful for educating clients
into the cognitive therapy perspective on anxiety. A number of other excellent self-help
manuals have been published as well on cognitive therapy or CBT for anxiety disorders
that can be given to clients as assigned reading. Appendix 6.1 presents a selected list of
self-help manuals that are consistent with the cognitive model. Often clients are even
more accepting of cognitive therapy after reading published accounts because it provides
external validation that cognitive therapy is a well established and widely recognized
treatment for anxiety.
Clinician Guideline 6.7
In the initial sessions of cognitive therapy, focus on educating the client into the cognitive
model of anxiety and providing a rationale for treatment. Describe clinical anxiety as an
automatic affective response to inappropriate fear activation that overtakes one’s mental
operating system. The goal of cognitive therapy is to deactivate, or “turn off,” the fear program through deliberate and effortful changes in how we think and respond to anxiety. Educate clients into the cognitive model not by minilectures but by emphasizing its applicability
to their personal experience of anxiety.
Self-Monitoring and the Identification of Anxious Thoughts
Teaching clients how to catch their anxious thoughts has been a central ingredient in
cognitive therapy for anxiety since its inception (Beck et al., 1985). And yet this is one of
the hardest skills for clients to master. The reason is that anxious thinking can be very
difficult to recall when the person is in a nonanxious state. However, when individuals
are highly anxious, they can be so overwhelmed with anxiety that any attempt to record
anxious thinking is practically impossible. Moreover, it is during periods of intense
anxiety that the person is most likely to exhibit the exaggerated estimates of threat
probability and severity that are the core cognitive basis of anxiety (Rachman, 2006).
Thus in cognitive therapy for anxiety considerable effort is focused on training in selfmonitoring automatic anxious thoughts. Rachman (2006) also notes that it is important
to identify the current threat that maintains anxiety. Daily diaries and self-monitoring
of anxiety will play a critical role in identifying the perceived threat in everyday life.
There are two ways to introduce anxious clients to thought recording. First, have clients focus on writing down anxiety-provoking situations, rating their anxiety level, and
noting any primary physical symptoms and any behavioral responses. These aspects of
198
ASSESSMENT AND INTERVENTION STRATEGIES
anxiety are often readily available to individuals, and will give them practice in tracking
and dissecting their anxiety episodes. Second, it is important that the first introduction
to anxious thinking be done in the therapy session (Beck et al., 1985). Since clients are
often not anxious while in session, some form of mild anxiety induction exercise may
be needed to elicit anxious thinking. For example, a panic induction exercise such as 2
minutes of overbreathing or spinning in a chair could be used to induce panic-like physical sensations. The client could be asked to verbalize any thoughts related to the exercise
such as fear of heart attack, fainting, losing control, or the like. A person with PTSD
could be asked to recall aspects of a past trauma and then verbalize his present thoughts
about his recollected memories. A mild fear of contamination or doubt could be induced
with someone suffering from OCD to elicit her appraisals about the threat. In each case
the therapist would ask probing questions about the client’s immediate thoughts. “What
went through your mind as you were breathing harder and harder?,” “What were your
thoughts focused on?”, “What were your main concerns?”, “What was the worst that
could happen?”, “Did it feel like the worst outcome was likely to happen?”, “Did you
have any competing thoughts, such as maybe it wasn’t so bad after all?”
Once the client has demonstrated some rudimentary skills at identifying his initial apprehensive (automatic) thoughts and appraisals in the therapy session, the therapist should assign a self-monitoring homework task. The Apprehensive Thoughts SelfMonitoring Form (Appendix 5.4) will be especially useful in this regard. Most clients
need extended practice in self-monitoring their anxious thoughts between sessions. In
fact self-monitoring of anxious thoughts and symptoms will continue throughout the
course of treatment. Cognitive restructuring and empirical hypothesis testing can not be
successfully employed until clients have become capable of identifying their automatic
threat-related thinking.
It is important that the self-monitoring component of treatment increase the anxious person’s awareness of two primary characteristics of anxious thinking:
•฀ Overestimated probability appraisals—“Am I exaggerating the likelihood that
some threat or danger will happen?”
•฀ Exaggerated severity appraisals—“Am I overly focused on the worst possible outcome? Am I exaggerating the severity of a negative outcome?”
Sensitizing clients to their evaluations of threat is important in shifting their focus from
threatening content (e.g., “What if the medical tests indicate cancer?”) to how their
appraisals contribute to anxiety (e.g., “Am I exaggerating the probability that the test
will be positive and lead to the worst possible outcome?, If so, what effect is this having on my anxiety?”). Individuals will need repeated practice in identifying their initial
apprehensive thoughts in order to improve their ability to catch the exaggerated threat
appraisals. When reviewing self-monitoring homework, the cognitive therapist probes
for exaggerated likelihood and severity of threat appraisals in order to reinforce the
importance of this thinking in the persistence of anxiety.
Homework Compliance
Homework compliance is an important issue in cognitive therapy for anxiety and often
it will be felt most keenly at the early phase of treatment when first assigning self-
Cognitive Interventions for Anxiety
199
monitoring homework. Many clients do not like filling in forms or writing about their
anxious thoughts and feelings. Even though there is mounting empirical evidence of
an association between treatment improvement and homework compliance (Kazantzis,
Deane, & Ronan, 2000), many clients still have great difficulty engaging in homework.
This problem has been addressed in a number of recent volumes on cognitive therapy,
and various suggestions have been offered for improving homework compliance (see
J. S. Beck, 2005; Leahy, 2001; Kazantzis & L’Abate, 2006). In the present context
the therapist should deal with any misconceptions or difficulties the client may have
about homework. The importance of homework and learning to identify anxious thinking should be emphasized as an essential skill that must be acquired before utilizing
the other cognitive and behavioral strategies for reducing anxiety. Homework should
be assigned in a collaborative fashion with instructions written for client convenience.
However, if an individual persists in refusing to engage in homework, termination of
further treatment may be necessary.
There is one reason for homework noncompliance that may be specific to the anxiety disorders. Sometimes clients are reluctant to engage in any self-monitoring of their
anxious thoughts and symptoms because they are concerned it will make the anxiety
worse. For example, a 33-year-old man with abhorrent obsessions about pedophilic
sex was afraid that writing down the occurrence and accompanying appraisals of the
thoughts would not only make them more frequent and raise his anxiety level, but these
thoughts were also a violation of his moral values. He was also concerned that drawing even more attention to the thoughts would erode what little control he had over
the obsessions. In this example concerns about escalating anxiety, the repugnant and
immoral nature of the obsessions, and fear of losing control all contributed to reluctance
to engage in self-monitoring his anxious thoughts.
A number of steps can be taken to address this situation. First, it is important to
make homework noncompliance a therapeutic issue. The faulty beliefs contributing to
reluctance to self-monitor anxious thoughts should be identified and cognitive restructuring can be utilized to examine these beliefs and generate alternative interpretations.
Possibly the homework assignment could be broken down into less threatening steps
such as asking the client to experiment with self-monitoring thoughts on a certain day
(or period within a day) and record the effects of the monitoring. This would be a direct
behavioral test of the belief that “writing down my anxious thoughts will make me
more anxious.”
The cognitive therapist should spend time during the educational phase introducing
the importance of homework and then periodically throughout treatment reminding the
client of the role that homework plays in the success of cognitive therapy. The following
is one way to explain homework to anxious clients:
“Homework assignments are a very important part of cognitive therapy. Approximately 10–15 minutes toward the end of each therapy session, I will suggest that
we summarize the main issues we’ve dealt with in the session and then decide on
a homework assignment. We will discuss the assignment together and make sure
it is something that you agree is doable. I will write the assignment down so we
are both clear on what needs to be done. From week to week I will also be giving
you different types of forms on which to record the results of the assignment. The
assignments will be short and not involve more than a few minutes out of your day.
200
ASSESSMENT AND INTERVENTION STRATEGIES
At the beginning of each session I will review last week’s homework with you. You
can expect that each week we’ll spend at least 10–15 minutes of the session reviewing the outcome of the homework and any problems you may have encountered. Do
you have any questions at this point?
“You may be wondering, do I really have to do homework? I always hated
homework in school. Besides I’m too busy for this sort of thing. You can think of
cognitive therapy like ‘mental exercise.’ In any physical training program, you need
to run, walk, or go to the gym three to five times a week in order to gain strength or
lose weight. You wouldn’t expect to meet your physical goals just by meeting with
the trainer once a week. The same thing happens in cognitive therapy. You are developing a different mental approach to your anxiety that involves learning to respond
to anxiety in ways that are not natural to you. You need lots of practice in using this
alternative approach to override the automatic anxiety program. Switching off the
anxiety program takes repeated practice and it won’t happen just by meeting with
the therapist once a week. The best way to overcome anxiety is through repeated
practice in your daily life so that gradually the new way of responding becomes
second nature to you. Just like in physical exercise, we’ve found in our research that
cognitive therapy is most effective for people who do homework. Very often when
clients do not benefit from treatment one of the main reasons is that they have not
been doing homework. How do you feel about this aspect of therapy? Are you able
to make a commitment to engage in homework at this time?”
Clinician Guideline 6.8
One of the first skills taught in cognitive therapy is the ability to identify and record the
automatic apprehensive thoughts, images, and appraisals that characterize anxious episodes. In addition clients write down their observations of the physical and behavioral
symptoms of anxiety. Self-monitoring anxious thoughts is a prerequisite skill for cognitive
restructuring. It may be necessary to deal with homework noncompliance at this point in
therapy.
Cognitive Restructuring
The goal of cognitive restructuring is to modify or literally “restructure” a person’s anxious beliefs and appraisals about threat. It is an integral part of treatment for deactivating the anxiety program. The focus is on “current threat,” that is, what is perceived as
dangerous or threatening at this moment. Also the cognitive restructuring interventions
are directed at the appraisals of threat rather than at threat content. The central question is “Am I exaggerating the probability and severity of threat and underestimating my
ability to cope?” and not whether a threat could happen or not. For example, in panic
disorder cognitive structuring would focus on whether the client is relying on exaggerated and biased appraisals of bodily sensations. The therapist would avoid any debate
on whether or not the client could have a heart attack. The same is true for social phobia
where the focus is on probability and severity appraisals of perceived negative evaluation
from others and not on whether some people may be having negative thoughts about
them. In this section we describe six cognitive intervention strategies: evidence gath-
Cognitive Interventions for Anxiety
201
ering, cost–benefit analysis, decatastrophizing, identifying cognitive errors, generating
alternatives, and empirical hypothesis testing.
Evidence Gathering
This intervention involves questioning clients on the evidence for and against their belief
that a threat is highly probable and will lead to severe consequences. Evidence gathering is the sine qua non of cognitive restructuring (Beck et al., 1979, 1985, 2005) and
has been variously labeled verbal disputation, logical persuasion, or verbal reattribution
(Wells, 1997). After identifying a core anxious thought or belief and obtaining a belief
rating on the thought, the therapist asks the following questions:
•฀ “At the time when you are most anxious, what is happening that convinces you
the threat is highly likely to occur? Is there any evidence to the contrary, that is,
that the threat is not likely to occur?”
•฀ “When you are feeling most anxious, what evidence is there that the outcome will
be so serious? Is there any contradictory evidence that the outcome may not be as
bad as you are thinking?”
•฀ “What makes the evidence for your anxious thinking believable?”
•฀ “Do you think you might be exaggerating the probability and severity of the
outcome?”
•฀ “Based on the evidence, what is a more realistic or likely estimate of the probability and severity of the worst that might happen?”
Appendix 6.2 provides an evidence-gathering form that can be used with clients.
The therapist and client first write down the primary anxious thought or belief that
characterizes an anxious episode. The client then provides probability and severity estimates based on how he feels during anxiety episodes. Using the Socratic form of questioning, the therapist probes for any evidence that supports such a high probability and
severity estimate of outcome. Although Appendix 6.2 is limited to six entries, additional
pages may be necessary to fully document the evidence supporting the anxious thought
or belief. After writing down all the supporting evidence, the therapist then asks for evidence that suggests the probability and severity estimates may be exaggerated. Normally
the therapist has to take more initiative in suggesting possible contradictory evidence
because anxious individuals often have difficulty seeing their anxiety from this perspective. Once all the evidence against the anxious thought or belief has been recorded the
client is asked to rerate the likelihood and severity of the outcome based solely on the
evidence.
Individuals will sometimes protest, saying “Yes, but when I’m anxious it feels like
the worst is going to happen even though I know it probably won’t happen.” The cognitive therapist should remind the client that “evidence gathering” is simply one approach
out of many that can be used to deactivate anxiety. Whenever the client feels anxious,
what has been learned from evidence gathering can be used to lower threat probability
and severity appraisals to a more realistic level, thereby countering a major factor in the
escalation of subjective anxiety. The following clinical example illustrates an evidencegathering approach with a 27-year-old traveling salesperson who suffered from panic
disorder and mild agoraphobic avoidance.
202
ASSESSMENT AND INTERVENTION STRATEGIES
THERAPIST: Renée, I notice from your panic log that last Wednesday you were driving
alone to a retail customer along a route that you don’t normally take when you suddenly felt like you couldn’t breathe. You indicated that you pulled the car to the side
of the road and got out to get some fresh air. You wrote down a number of bodily
sensations like a lump in your throat, feeling like you couldn’t get enough air, grasping for breath, chest tightness, heart palpitations, dizziness, and general tension.
CLIENT: Yes, it was one of the worst panic attacks I’ve had in a long time. I just couldn’t
seem to breathe properly. The harder I tried the worse it got. I took sips of water to
clear my throat but that didn’t help.
THERAPIST: What were your afraid might be happening?
CLIENT: I was really afraid that I would suffocate. That’s what it felt like. Here I was
alone, in the middle of nowhere, and I couldn’t breathe. It got so bad I wondered if
I could actually suffocate to death.
THERAPIST: Okay, Renée, let’s write that anxious thought—“Thought that I would suffocate alone and die”—here on this line using a form called “Testing Anxious
Appraisals: Looking for Evidence” (Appendix 6.2). Now I would like you to think
back to when you had the panic attack. When you were off to the side of the road,
alone, and struggling to catch your breath, how likely was it that you were suffocating to death? In other words, based on how you were feeling, what did the probability that you were suffocating feel like to you from 0% (no chance it will happen)
to 100% (certain it is happening).
CLIENT: Well, at the time it felt like a 90% probability that I was suffocating.
THERAPIST: And what about the severity of the outcome? How serious did it feel to
you? Were you focused on the worst possible outcome like death by suffocation or
something less serious like feeling the discomfort of panic? What rating would you
use from 0 to 100 to indicate how serious the consequence seemed to you when you
were panicky.
CLIENT: Oh, it was serious. All I could think about was suffocating to death by myself.
I would give this a 100 rating.
THERAPIST: Okay, now let’s look at the evidence, such as anything that was happening
at the time, or past experiences, or information of any kind that would indicate that
you were at high risk of experiencing a serious outcome like death by suffocation.
CLIENT: Well, one thing that makes me wonder if this suffocation feeling is serious
is that it comes on me so suddenly, out of the blue. One minute I’m fine and then
before I know it I can’t breathe.
THERAPIST: Okay, let’s put that down on the first line under evidence for the anxious
thought: “onset of suffocation feeling is rapid and unexpected.” Any other evidence that makes you think you’re likely to suffocate to death?
CLIENT: The anxiety associated with this feeling is very intense, even panic. It’s so bad
I’m convinced something serious must be happening.
THERAPIST: Let’s put that down under the second entry: “feel intensely anxious, even
panicky.” Anything else?
CLIENT: Well, the fact that I try to calm myself down by taking long, slow breaths and
Cognitive Interventions for Anxiety
203
yet it doesn’t help makes me convinced something is terribly wrong. If all of this was
just bad nerves, shouldn’t it go away when I breathe more slowly?
THERAPIST: Okay, a third piece of evidence for the anxious thinking is “controlled
breathing doesn’t make the suffocation feelings disappear.” Is there anything else?
CLIENT: As I mentioned previously, I have very vivid memories of my uncle grasping for
breath. He had a long battle with emphysema which in the final stage of the disease
left him unable to breathe. It was a most horrifying way to die.
THERAPIST: So the fourth evidence for the anxious thought is “memories of an uncle
who eventually died from suffocation because of emphysema.” Is there any other
evidence for your anxious thoughts about suffocation?
CLIENT: No, that pretty well covers it.
THERAPIST: Now let’s look for evidence that does not support the view that you are at
high risk of dying from suffocation. Can you think of any?
CLIENT: This is much harder to do. Huh. . . . I suppose one thing is that I haven’t died yet.
I mean I’ve had these suffocation feelings for months and yet I’m still here.
THERAPIST: Have you come close to death? For example, did you ever almost pass out,
turn blue, or were paramedics ever called to provide you with oxygen?
CLIENT: No nothing like that. I’ve never had any tangible signs that I’m dying from suffocation.
THERAPIST: Let’s write that down as evidence against your anxious thought on this first
line in the right-hand column of the form: “I’ve never experienced any tangible
medical signs that I am dying from lack of oxygen.” Can you think of anything
else?
CLIENT: Well, my family doctor has ordered various medical tests and I have seen specialists but they all say I am healthy. They say my respiration system is fine.
THERAPIST: So a second piece of evidence against the anxious thought is that “I am
physically very healthy as far as can be determined by medical science.” Is there
any other evidence?
CLIENT: I can’t think of any.
THERAPIST: Well, how hard is it to stop breathing? How long can you hold your breath?
Let’s try it. [Therapist times client on length of breath holding.]
CLIENT: That was really hard, even though I tried not to breathe, eventually I couldn’t
help myself. I had to breathe.
THERAPIST: Exactly, breathing is an automatic response. It is very difficult to stop
breathing, even when you try your very best. Because breathing is such an automatic physiological response, people rarely just stop breathing spontaneously for no
apparent reason. Have you ever heard of that happening to someone?
CLIENT: No, I haven’t.
THERAPIST: So let’s write that down as the third piece of evidence against your anxious
suffocation thought: “Breathing is such an automatic physiological response it is
exceedingly rare to suddenly, unexpectedly stop breathing for no apparent reason.” Can you think of any other contrary evidence?
204
ASSESSMENT AND INTERVENTION STRATEGIES
CLIENT: No, I’m stuck.
THERAPIST: Have you ever noticed whether there is anything you can do that reduces the
suffocation feelings? For example, what happens to the feelings if you get distracted
or are busy at work?
CLIENT: Well, on a couple of occasions when I started to get the feeling of not catching
my breath and then I got real busy at work, I somehow forgot about it and the feelings went away.
THERAPIST: Okay, so maybe distraction can cause a reduction in suffocation feelings. Is
there anything that seems to make the feelings worse?
CLIENT: My worst panic attacks have been when I’m driving alone in the car along a
remote, unfamiliar highway. I seem to become really preoccupied with my physical
state.
THERAPIST: Is there any chance, then, that focusing on breathing sensations makes the
suffocation feelings worse?
CLIENT: It is possible.
THERAPIST: So let’s write this down as the fourth evidence against the anxious thought:
“Suffocation feelings are worse when I focus on my breathing and least when I
am distracted.” Does that sound like a condition that could lead to death? Do you
suppose doctors warn people not to focus on their breathing because it might cause
them to suffocate or if they have breathing problems, just distract themselves? Does
this sound like a cure for emphysema?
CLIENT: No, obviously not. But I suppose it is consistent with anxiety as the cause of
suffocation feelings. This is what my doctors have been telling me.
THERAPIST: Okay, so let’s rerate your anxious thought “I will suffocate alone and die.”
Based on the evidence (and not on your feelings), what is the likelihood that you
will die from suffocation?
CLIENT: Well, I suppose it is much less than 90% but it is certainly not zero. I’ll say
20%.
THERAPIST: And based solely on the evidence, how serious is the likely outcome of your
suffocation feelings?
CLIENT: Again, it’s probably not 100% because death is highly unlikely. I guess the
seriousness is about 60%.
THERAPIST: What this tells us is that you tend to overestimate the probability and severity of threat (“I’ll die from suffocation”) when you are anxious. However, when
you focus on the evidence (and not on your feelings) you realize the threat is much
less severe. We know that making exaggerated threat estimates makes anxiety
worse and when a person sees the threat more realistically, her anxiety declines.
So, one way to reduce your anxiety is to correct your exaggerated anxious thoughts
by reminding yourself of all the evidence against the thought and then rerating its
probability and severity. After you do this a few times using Appendix 6.2 as a
guide you’ll become skilled at correcting your anxious thinking.
CLIENT: That sounds fine but when I’m really anxious I can’t think straight.
Cognitive Interventions for Anxiety
205
THERAPIST: I understand but the more you practice correcting the anxious thoughts and
appraisals, the more automatic the whole process becomes and the better you will
get at using this technique to reduce your anxiety. Would you like to give this a try
with a homework assignment?
CLIENT: Sure, let’s give it a try.
[Note: If the client had difficulty utilizing evidence gathering in response to in vivo anxiety episodes, the therapist could have the client imagine panic situations and practice
countering the anxiety with contradictory evidence.]
Clinician Guideline 6.9
Teach clients how to gather evidence for and against their appraisals of heightened probability and severity of threat related to their anxious concerns. Threat probability and severity
estimates are recalculated solely on the basis of the evidence that is generated. Evidence gathering can be an effective method of challenging exaggerated anxious thinking by encouraging the anxious person to shift from affect-based appraisals (i.e., ex-consequentia reasoning:
“I feel anxious, therefore I must be in danger”) to evidence-based appraisals of a situation.
Cost– Benefit Analysis
In anxiety disorders cost–benefit analysis is a particularly versatile and effective intervention because individuals are already focused on the consequences of their thoughts
and feelings. The therapist helps the client consider the question “What is the consequence, the advantages and disadvantages, of holding this particular belief or perspective in regards to my anxiety?” (see Leahy, 2003). Wells (1997) noted that cost–benefit
analysis can also improve motivation for treatment. After identifying a core anxious
thought, belief, or appraisal, the therapist can pose the following questions:
•฀ “From your experience, what are the immediate and long-term consequences of
embracing this anxious thought?”
•฀ “Are there costs and benefits, or advantages and disadvantages to believing in the
anxious thought?”
•฀ “What immediate and long-term effect does this thinking have on your anxiety?”
•฀ “If you had a different outlook on your anxiety, what would be the costs and
benefits?”
The therapist can use the Cost– Benefit Form in Appendix 6.3 to conduct a cost–
benefit analysis on an anxious thought or belief. The anxious thought is first recorded.
Then, using guided discovery, the therapist explores the immediate and long-term
advantages and disadvantages of accepting the anxious thought. Clients are asked to
circle the consequences, both positive and negative, that are really important to them.
Next an alternative way of thinking about the anxious situation is considered and the
206
ASSESSMENT AND INTERVENTION STRATEGIES
costs and benefits of this approach are written in the lower half of the form. Again the
consequences of most significance to the client are circled. The objective of this exercise
is to emphasize the heavy costs associated with anxious thinking and the immediate
benefits of an alternative perspective. Homework assignments can be constructed to test
out the consequences of anxious thinking and the benefits of an alternative approach.
The therapist encourages clients to practice shifting their focus when anxious from
the threat content to the question “Is this anxious thinking helpful or harmful?” The
therapist emphasizes that repeatedly reminding oneself of the costs of anxious thinking
is another effective way to weaken or deactivate the anxiety program. This intervention
is particularly effective if clients fully realize that anxious thinking actually fuels their
discomfort rather than helping them cope with or avert the perceived threat.
Jeremy suffered for years with GAD. One of his primary worries concerned
finances despite having secure, well-paid employment. He always paid his bills, had
enough money each month to contribute to his investment account, and had never even
approached bankruptcy or financial hardship. And yet Jeremy continually worried that
he was not putting enough into his investments and as a result he would not be prepared
for the possibility of financial ruin. Using the Cost– Benefit Form, we identified Jeremy’s
anxious thought as “I’m not saving enough money to prepare myself for the possibility
of some future financial disaster” (e.g., losing my job and having no income). Jeremy
believed that his worry about saving money had a number of important advantages
such as (1) it forced him to save more each month and so his investments were growing,
(2) he watched his expenses much more closely, (3) he’ll be better prepared to absorb a
financial loss, (4) it ensures that he wouldn’t lose the house or go bankrupt if he did lose
his job, and (5) he feels better about himself when he is saving. Jeremy circled (1) and (3)
as the most significant advantages of his worry.
Exploring the disadvantages of the worry proved more difficult but with therapist guidance the following list was generated: (1) the more he thinks about not saving
enough, the more anxious and tense he feels; (2) once he starts to worry about saving
enough, he can’t seem to stop it, it completely takes over his mind; (3) he hasn’t slept
well because of worry over his savings; (4) there is little enjoyment in his life because
he is constantly worried about finances; (5) he frequently deprives himself of little pleasures for fear of spending money; (6) he gets into severe arguments with his wife over
saving and spending money and she has threatened to leave; (7) he feels distant and
uninvolved with his children because of his preoccupation with finances; and (8) he
spends long, frustrating hours each night monitoring his investments. Jeremy indicated
that (1), (3), (6) and (7) were the most important costs associated with his worry over
saving money.
At this point in therapy, Jeremy was still heavily invested in his financial worries.
As a result he had difficulty generating an alternative perspective. Eventually, after considerable discussion, it was agreed that the following way of thinking could become a
goal of treatment. “I am saving enough money for temporary, moderate financial loss
but there is little I can do to guarantee protection against a sustained period of total
financial ruin.” We then discussed a number of advantages to this perspective on his
finances: (1) less anxiety about saving because he no longer needs to amass a huge safety
net of savings, (2) more tolerance for stock market fluctuations, (3) less need to monitor
his investments, (4) more freedom to spend on everyday pleasures and comforts, and
(5) fewer conflicts with spouse over finances because of less attempt to control expen-
Cognitive Interventions for Anxiety
207
ditures. Both (1) and (5) were marked as significant advantages of the new perspective. In terms of disadvantages, Jeremy wondered (1) if he might end up with a smaller
investment account because he is saving less money, and (2) he would be prepared for a
narrower range of future financial losses. Overall, Jeremy agreed that the disadvantages
of the anxious thoughts about saving and the advantages of adopting a more moderate
view were clearly evident.
The therapist was able to use the cost–benefit analysis in future sessions by reminding Jeremy to think about “the consequences of anxious thinking about savings” and
the benefits “of thinking about moderate savings.” In particular, whenever Jeremy
engaged in anxious worry about his finances, the therapist reminded him to “remember
the cost–benefit form, and what you are doing to yourself by worrying about saving for
the ultimate financial disaster.” “Based on the cost–benefit analysis, how can you think
about saving that will lead to less anxiety?” Again clients are reminded that repeatedly
thinking in terms of cost–benefit analysis whenever they engage in anxious thinking is a
useful tool for weakening the fear program and diminishing their anxiety.
Clinician Guideline 6.10
Cost–benefit analysis is a cognitive intervention that teaches clients to take a pragmatic
approach by examining the immediate and long-term advantages and disadvantages of
assuming exaggerated threat, or alternatively, of adopting a more realistic perspective. The
therapist uses guided discovery and homework assignments to help clients achieve a full
realization of the heavy costs associated with “assuming the worst” and the benefits derived
from a more realistic alternative perspective. Clients can use this insight to counter their
anxious thoughts and beliefs.
Decatastophizing
A third cognitive intervention that can be especially useful for most anxiety disorders
involves having the client “hypothetically” confront his dreaded catastrophe or the
worst that could happen. Beck et al. (1985, 2005) provided an extensive discussion
of the use of decatastrophizing to modify exaggerated threat appraisals and beliefs.
Craske and Barlow (2006) describe castatrophizing as “blowing things out of proportion” (p. 86) and decatastrophizing as “imaging the worst possible outcome and then
objectively judging its severity” (p. 87). They note that catastrophizing involves thinking
about outcomes that are entirely unlikely to happen, even impossible (e.g., “I could get
a mental illness by coming in close proximity to a homeless person”), or exaggerating
events that are highly unlikely (e.g., “People will notice I am nervous and think I am
mentally unstable”), or jumping to an extreme conclusion from a minor event (e.g., “If
I make a mistake on this form, it will be completely invalidated and I won’t receive my
long-term disability benefits”).
Decatastrophization is an effective intervention when it is clear that catastrophic
thinking is apparent in the client’s threat and vulnerability appraisals. It is a particularly useful approach when dealing with the cognitive avoidance evident in pathological worry (Borkovec et al., 2004). Decatastrophizing confronts cognitive avoidance by
encouraging the client to face the imagined catastrophe and its associated anxiety. This
208
ASSESSMENT AND INTERVENTION STRATEGIES
intervention also has a number of other therapeutic benefits such as realigning threat
probability and severity estimates to a more realistic level, increasing a sense of selfefficacy for dealing with future negative outcomes, and enhancing information processing of safety and rescue features in future dreaded situations.
There are three components to decatastrophizing:
1. Preparation stage
2. Description of catastrophe (“What’s the worst that could happen?”; “What
would be so bad about that?”)
3. Problem-solving stage
Timing is everything when using decatastrophizing. Given the level of anxiety and avoidance often associated with “thinking about the worst-case scenario,” other cognitive
and behavioral interventions should be employed as preparation for this form of “imaginal exposure.” Later in the course of therapy, decatastrophizing could be introduced
as a way of confronting “the fears in your own mind.” The rationale and benefits of the
intervention should be explained and the client’s readiness to engage in decatastrophizing should be evaluated.
Assuming proper timing and preparation, the next step is to obtain a complete,
detailed discussion of the worst-case scenario from the client. Probing questions such as
“What’s the worst that can happen” or “What’s so bad about that” can be used. The
downward arrow technique is often useful for arriving at the dreaded catastrophe. The
client should be encouraged to describe all aspects of the feared catastrophe including
its consequences to self and others (“How would your life change?”), its probability of
occurrence, its severity, and the client’s perceived inability to cope. The therapist should
determine whether the client recognizes any safety or rescue features in the worst-case
scenario. If possible, imaging the catastrophe is a more potent way to obtain the emotionally charged aspects of the worst possible outcome. Ratings on anxiety experienced
while discussing or imagining the catastrophe should be obtained as a way of demonstrating the anxiety-inducing effects of catastrophizing. Having the client provide a
written description of the catastrophe is an effective way to reduce possible cognitive
avoidance that can occur when imagining or even discussing the “worst-case scenario.”
Also the therapist should determine the client’s level of insight into the exaggerated or
irrational nature of the catastrophizing and its effects on anxiety.
After a clear description of the worst-case scenario, a problem-solving approach
to catastrophizing can be introduced as a way to counter this form of thinking. The
evidence-gathering approach can be used to evaluate the likelihood of the worst-case
scenario. A best possible outcome can also be developed as a way of framing the most
extreme negative and positive outcome (Leahy, 2005). A more realistic, middle-of-theroad, negative outcome can be developed as an alternative to the catastrophe. Together
the therapist and client can work out an action plan that would involve how to cope
with the more realistic negative outcome. This action plan would be written out and the
client encouraged to work on the plan whenever she started to catastrophize.
As an example, Josie had two to three severe panic attacks on a daily basis. She was
hypervigilant about her pulse rate and became very anxious whenever she perceived that
her heartbeat was too fast and possibly irregular. She engaged in catastrophic misinterpretation of her pulse rate, believing that she would have a life-threatening heart attack
Cognitive Interventions for Anxiety
209
if her heart rate was too high. After a number of sessions involving education and less
threatening interventions that focused on her misinterpretation of an accelerated heart
rate, Josie agreed to engage in an imaginal exposure exercise in which she imagined an
accelerated pulse rate that led to a severe heart attack in which she is lying on the ground
grasping her chest, alone and dying. After an extensive evidence-gathering exercise in
which the client and the therapist were able to examine evidence on whether the probability and severity of the catastrophic outcome was exaggerated or not, an alternative
negative outcome was considered. Josie decided that a more likely very negative outcome
might be chest tightness, a racing heart, followed by severe chest pain. She rushes to an
emergency room and finds out she has had a mild heart attack. An action plan was then
developed on how she would manage the rest of her life knowing that she has a heart
condition. Josie was encouraged to work on the “mild heart attack” plan whenever
she started to catastrophize. Notice that the purpose of this intervention was to reduce
Josie’s fear (and catastrophizing) about heart disease and to increase her perceived ability to cope if this situation ever occurred.
Clinician Guideline 6.11
Decatastrophizing involves the identification of the “worst-case scenario” associated with an
anxious concern, the evaluation of the likelihood of this scenario, and then the construction
of a more likely moderate distressing outcome. Problem solving is used to develop a plan for
dealing with the more probable negative outcome.
Identifying Thinking Errors
Teaching anxious clients to become more aware of the cognitive errors they tend to make
when feeling anxious is another useful strategy in the modification of faulty appraisals of anxiety. Highlighting the errors in one’s thinking style reinforces the message to
clients that threat perceptions are inaccurate when people are highly anxious. It encourages a more critical, questioning approach to one’s anxious thinking. Thus it is important that clients understand the rationale for identifying and then correcting cognitive
errors. The therapist could use the following explanation:
“Although everyone engages in these erroneous thinking styles from time to time,
these errors are particularly prominent when we are anxious. When we commit
these errors in our thinking they tend to lead to more exaggerated and biased conclusions. For example, if I always focused only on the flaws or mistakes in a talk
whenever I gave a public address (tunnel vision error), I would end up concluding
that the talk was terrible and I was a dismal failure. The same thing happens when
we commit these cognitive errors when we’re feeling anxious. They lead us to exaggerated and false conclusions about the threat or danger in a situation and our
inability to cope. So learning to identify these errors and correct them is an important intervention for reducing anxious thoughts and feelings.”
Clients can be given a copy of Appendix 5.6 in order to familiarize themselves with
the six forms of cognitive errors that are common in anxiety: catastrophizing, jumping
210
ASSESSMENT AND INTERVENTION STRATEGIES
to conclusions, tunnel vision, nearsightedness, emotional reasoning, and all-or-nothing
thinking. Error identification should be introduced by first going over thought records
produced in the session and discussing cognitive errors that are apparent in the client’s
anxious thinking. This can be followed with a homework assignment in which clients
record examples of thinking errors taken from their everyday experiences (use Appendix
5.6). After this exercise the therapist encourages clients to incorporate error identification into a cognitive strategy utilized whenever they engage in anxious or worrisome
thinking.
Taylor (2000) describes an inductive reasoning approach that can be very useful
in countering the erroneous thinking style that leads to exaggerated threat appraisals.
Through the use of Socratic questioning and a guided discovery approach, the therapist
explores with the client how a particular situation or symptom can lead to a dreaded
outcome. For example, a client could be asked how tightness in the chest could cause a
heart attack, or how lying down prevents such heart attacks. A person with PTSD who
becomes anxious when remembering a past trauma could be asked how such recollections increase the likelihood of present danger or a future trauma. Individuals with
repugnant sexual obsessions could be asked to explain how such thoughts would lead
to committing a sexual offense, or a person with social phobia could explain how a
nervous feeling would lead to public humiliation. By engaging in this form of inductive
questioning, the therapist is provided material that can be used to highlight the cognitive errors in anxious thinking that lead to faulty conclusions about threat and personal
vulnerability.
Clinician Guideline 6.12
Clients learn to identify the cognitive errors and faulty inductive reasoning that characterizes an anxious thinking style. This intervention helps clients develop a more critical stance
toward their automatic anxious thoughts.
Generating an Alternative Explanation
During periods of heightened anxiety, an individual’s thinking is often extremely rigid and
inflexible, with a narrow focus on the perceived threat or danger (Beck et al., 1985, 2005).
Clients will often recognize that their anxious thinking is irrational but the strong emotional charge associated with the thoughts makes them difficult to ignore. Thus searching
for alternative explanations for anxious situations can be extremely difficult. Repeated
practice with the cognitive therapist coaching the client in generating alternative explanations to a variety of anxious situations will be necessary before this skill generalizes to
naturalistic anxious situations that occur outside the therapy setting. It may be necessary
to present the alternative as a tentative possibility that the client is encouraged to at least
entertain as another way to understand a situation (Rouf, Fennell, Westbrook, Cooper,
& Bennett-Levy, 2004). At the same time, learning to produce less anxious alternative
interpretations is a critical component of cognitive therapy for anxiety because clients
need credible explanations that replace their catastrophic interpretation.
The Alternative Interpretations Form in Appendix 6.4 can be used as a withinsession therapeutic tool or a homework assignment for generating alternative explana-
Cognitive Interventions for Anxiety
211
tions. Most clients will be able to produce the “most dreaded outcome” and the “most
desired outcome” with little difficulty. The alternative, more realistic or probable outcome will require considerably more prompting and guidance from the therapist. A
good alternative view should have the following characteristics:
1. Be clearly distinct from the catastrophic interpretation.
2. Have a better fit with the facts and reality of the situation.
3. Be amenable to empirical evaluation.
A client with OCD described as his primary obsession various disturbing sexual
thoughts about being gay. Although he was embarrassed by his apparent homophobic
reactions, nevertheless he continued to feel intensely anxious whenever situations triggered questioning thoughts about his sexual orientation. His catastrophic interpretation was “What if these frequent thoughts about being gay means that I am a latent
homosexual. I will then have to divorce my wife and move in with a gay lover.” His
most desired outcome was “Never to have thoughts about being gay and have absolute
certainty that I am 100% heterosexual.” The more probable alternative explanation was
“My frequent thoughts about being gay are not due to some latent homosexual orientation but rather to my overreaction to these thoughts because the thoughts represent a
violation of my personal moral standards.” Notice that the alternative interpretation is
a polar opposite to the catastrophic explanation. Whereas the anxious view is “These
thoughts may be caused by an unconscious homosexual orientation,” the alternative
explanation is “These thoughts are caused by a faulty response that stems from an
extreme aversion to a homosexual orientation (i.e., homophobia).” Evidence gathering
and empirical hypothesis testing are more effective when the alternative view and the
catastrophic explanation are polar opposites. In this way the results from such exercises
will be incontrovertible evidence for the alternative and against the catastrophic conclusion.
Table 6.3 presents examples of catastrophic interpretations, desired outcomes, and
alternative explanations that may be found in specific anxiety disorders. The goal is to
work with the client in generating credible alternative explanations that are subjected
to empirical verification. With practice the client can learn to replace the catastrophic
interpretation with the alternative explanation, thereby reducing the exaggerated threat
appraisals and associated subjective anxiety.
Clinician Guideline 6.13
Remediation of anxious thinking requires the discovery of more realistic alternative interpretations that can replace exaggerated threat-related appraisals. The most effective alternatives for countering automatic anxious thoughts and beliefs are those that offer a more balanced, evidence-based perspective that is clearly distinct from the anxious schemas.
Empirical Hypothesis-Testing
One of the most important interventions for cognitive change is behavioral experimentation or empirical hypothesis testing. First introduced in the cognitive therapy manual
212
ASSESSMENT AND INTERVENTION STRATEGIES
TABLE 6.3. Illustrations of Clients’ Catastrophic, Most Desired, and Alternative Explanations That May Be
Relevant for the Various Anxiety Disorders
Anxiety disorder
Catastrophic interpretation
Desired outcome
Alternative interpretation
Panic disorder
(lightheaded,
feelings of unreality)
“I’m losing control, contact
with reality. Maybe I am
going crazy and will have to
be hospitalized.”
“I want to always feel
fully conscious and
aware at any moment.”
“Feelings of unreality and
lightheadedness reflect normal
variations in arousal level that
can be affected by a variety of
internal and external factors.”
Social phobia
(observes signs of
increasing anxiety)
“Everyone will notice that
I’m getting anxious and
wonder what is wrong with
me. I’ll end up making a
fool of myself.”
“I want to always feel
perfectly relaxed and
confident in social
settings.”
“One can feel anxious and
still perform competently in a
social setting. Whether others
observe my anxiety and draw
negative conclusions can not
be known.”
Generalized anxiety
disorder (worry
about finishing
minor daily tasks)
“I will be so worried about
doing chores and errands
that I’ll be completely
paralyzed and have to be
re-hospitalized.”
“I want full confidence
and certainty that I
will accomplish the
daily goals that I set for
myself.”
“Worry will slow me down
and reduce the amount that
I can get done but it doesn’t
have to lead to complete
paralysis and inactivity.”
Obsessive–
compulsive disorder
(fear of mental
contamination or
morphing)
“If I get too close to people
I feel are weird or different,
I will lose my creative
potential.”
“I prefer to avoid all
contact with people
who are different and
threaten my creativity.”
“My creativity has been
hampered by my OCD rather
than by close proximity
to people I perceive as
undesirable.”
Posttraumatic stress
disorder (reaction to
recurrent assaultrelated images)
“My inability to suppress
these images means the
PTSD is so bad I will never
be able to function in life.”
“I desire no unwanted
recollections or
memories of the brutal
assault.”
“Everyone who has been
assaulted has to live with
disturbing memories while
minimizing their impact on
daily living.”
for depression (Beck et al., 1979), behavioral experiments are planned, structured experiences designed to provide the client with experiential data for and against threat and
vulnerability appraisals or beliefs. The Oxford Guide to Behavioural Experiments in
Cognitive Therapy, the most comprehensive clinical guide to behavioral experimentation, offered the following operational definition: “Behavioural experiments are planned
experiential activities, based on experimentation or observation, which are undertaken
by patients in or between cognitive therapy sessions” (Bennett-Levy et al., 2004, p. 8).
They are derived from a cognitive formulation of anxiety, and their main purpose is
to provide new information that can test the validity of dysfunctional beliefs, reinforce more adaptive beliefs, and verify the cognitive formulation. Based on conceptual
considerations, clinical experience, and some empirical evidence, Bennett-Levy et al.
(2004) make a compelling case for behavioral experimentation as the most powerful
therapeutic strategy available to cognitive therapists for promoting cognitive, affective,
and behavioral change.
In the anxiety disorders empirical hypothesis testing usually involves some form
of exposure to a fear situation and a disconfirmatory manipulation that tests the validity of the anxious appraisal (D. M. Clark, 1986b; Wells, 1997). The most effective
hypothesis-testing exercises are structured so that the outcome of the experiment can
Cognitive Interventions for Anxiety
213
refute the anxious belief and supports the alternative interpretation. Given the overwhelming empirical evidence for the effectiveness of exposure in fear reduction (see discussion in Chapter 7), exposure-based hypothesis-testing exercises are a key intervention in cognitive therapy of anxiety. Behavioral experiments should be introduced early
and continued throughout the course of treatment. Often they play a defining role in the
modification of anxious thinking. In fact it is difficult to imagine an effective cognitive
intervention for anxiety that does not include within- and between-session behavioral
exercises. Behavioral exercises can take the form of fairly spontaneous within-session
demonstrations such as asking the client to suppress thoughts of a white bear in order
to illustrate the negative effects of intentional thought suppression. In the following section we discuss the critical steps in developing an effective empirical hypothesis-testing
exercise. (See Rouf et al., 2004, for more detailed discussion of how to construct effective behavioral experiments.)
Step 1. The Rationale
Any empirical hypothesis-testing exercise should be derived from the primary issue of
the therapy session and it should be consistent with the cognitive case formulation. The
cognitive therapist introduces the exercise by providing a rationale. This can be illustrated in the following case example. Jodie was a 22-year-old university student who
developed an incapacitating anxiety about attending large lecture-based classes. Her
primary anxious thought was “Everyone in the class notices me and thinks that I don’t
belong in university.” This led to escape (ie., leaving class early) and avoidance (i.e.,
skipping classes) behaviors that were jeopardizing her academic performance. In this
situation the therapist introduced a behavioral experiment by stating:
“So, Jodie, you are sitting in class and feeling very anxious. You have the thought
‘everyone is probably looking at me and thinking she doesn’t belong in university.’
I wonder if we could come up with an experiment or some sort of exercise to test
the accuracy of this thought. I could ask you to try and remember reasons why
you think this interpretation might be true or false, but the most accurate way to
find out is to collect information on site. The very best way to test out this anxious
thought is to collect information on it while you are in the classroom. We all learn
so much more from our own experiences than we do from listening to teachers or
even therapists for that matter. In fact homework exercises such as this have been
shown to be one of the most important ingredients for reducing anxiety. Not only
does it give you an opportunity to test the anxious thinking, but it also provides
an opportunity for you to directly work on the anxiety. Would you like to work
together on constructing an exercise that would test out this anxious thought?”
Step 2. Statement of Threat Appraisal and Its Alternative
Assuming collaboration has been established with the client, the next step is to state the
threat appraisal and its alternative. The Empirical Hypothesis-Testing Form in Appendix 6.5 can be used to formulate the behavioral experiment and collect the outcome
data. A clear, specific statement of the threat interpretation (i.e., anxious thought or
belief) targeted by the exercise is essential for an effective behavioral experiment. The
214
ASSESSMENT AND INTERVENTION STRATEGIES
therapist should record the anxious appraisal on the Empirical Hypothesis-Testing Form
and ask clients to rate their belief in the statement on a 0–100 scale when they first
begin the behavioral exercise. The therapist and client then come up with an alternative
interpretation that is clearly distinct and more plausible than the anxious thought or
belief (see previous section on generating alternatives). The alternative is recorded on the
form and the client is asked to provide a belief rating at the conclusion of the behavioral
experiment. The two belief ratings will provide an indication of whether the behavioral
experiment has led to a shift in belief from a threat-related interpretation to the alternative perspective.
In our case illustration, Jodie’s threat interpretation was “If I feel nervous in class
everyone will notice me and think I don’t belong in university.” The alternative interpretation was “My nervous feelings are very evident to me but barely visible to my
classmates. Besides they are too busy listening to the lecture, talking to the person
beside them, sleeping, or daydreaming to take the time to notice me.” Each of these
statements was developed collaboratively during the therapy session for the behavioral
experiment.
Step 3. Planning the Experiment
Devising a good behavioral experiment will probably take at least 10–15 minutes of
therapy time. It is important to write out sufficient details of how the experiment should
be conducted so it is clear to the client what is to be done at a certain time and in a
particular location. The experiment must involve an activity that provides a clear test
between the anxious and alternative interpretation. It is important that the exercise is
planned out collaboratively with the client and there is agreement that the experiment
is a relevant test of the anxious thought. There is little sense in pursuing an empirical
hypothesis-testing exercise that the client doubts has relevance or has little intention of
carrying out. Assuming a mutually agreed-upon relevant exercise, the therapist should
write down specific instructions for completing the experiment in the left-hand column
of the Empirical Hypothesis-Testing Form.
Rouf et al. (2004) discuss a number of considerations that should be taken into
account when planning behavioral experiments. Make sure the purpose of the experiment is clear, that a time and place for the experiment has been identified, and that
resources needed to carry out the exercise have been determined. Any anticipated problems should be worked out prior to assigning the exercise. The therapist can ask a client “What do you think might discourage or even prevent you from carrying out this
exercise?” Problems such as insufficient time, limited opportunity, or heightened anticipatory anxiety must be addressed before assigning the exercise. It is important that
something constructive is gained from the experiment regardless of the outcome (i.e., a
win–win situation) and that the exercise is not too difficult or challenging for the client.
Finally all doubts, fears, and other concerns expressed by the client must be addressed
and any potential medical complications should be assessed by the client’s physician.
In our case illustration, the following behavioral experiment was constructed. Jodie
agreed to attend her next Chemistry 101 class on Wednesday at 9:00 A.M. She was asked
to arrive at the lecture hall at 8:55 and to sit at least three seats in from the aisle in a middle row. Ten minutes into the lecture she agreed to write down anything she noticed in
Cognitive Interventions for Anxiety
215
other students that indicated they were looking directly at her. Fifteen minutes into the
lecture she would take three to four deep breaths and observe whether anyone noticed
what she was doing. Twenty minutes into the lecture she would try to make her body
shake ever so slightly for a few seconds and observe whether anyone noticed. The therapist and client practiced each of the elements of the experiment: how to record student
reactions and what behavior would constitute a direct look, how to deep breath, and
how to shake ever so slightly. Jodie agreed that this was a “doable exercise” and that it
would be a good test of how much she is noticed in class.
Step 4. Hypothesis Statement
Under item 3 on the Empirical Hypothesis-Testing Form (Appendix 6.5), a specific
hypothesis can be recorded that reflects the client’s predicted outcome of the experiment. The hypothesis would directly reflect the anxious thought or belief stated in item
1. The therapist can ask, “Based on your anxious thought [state item #1 here], what do
think will happen when you do this exercise? What outcome would make you feel more
anxious?” In the present case Jodie wrote the following hypothesis on the form “Anything I do in class that is out of the ordinary such as arrive just before the beginning of
the lecture, take deep breaths, or slightly shake will draw attention to myself. Once I
notice people looking at me, I will feel intensely anxious.” Notice that the hypothesis
is derived from the anxious interpretation (i.e., “if I feel nervous in class everyone will
notice me and think I don’t belong in university”) but it is a more specific application of
the threat interpretation to the actual experiment.
Step 5. Record the Actual Experiment and Outcome
Clients should record how they conducted the experiment and its outcome as soon after
completing the exercise as possible. A short description of what was done and its outcome can be written in the center and right columns on the Empirical HypothesisTesting Form. Often individuals do not conduct an experiment exactly as planned so
a description of what was actually done is important in evaluating the success of the
exercise. However, the actual outcome reported by the client is even more important
when following up on the effects of the behavioral experiment. It is the client’s perceived
outcome that will provide the necessary information for determining whether the exercise had an effect on anxious thoughts and feelings. Thus the outcome recorded on the
form becomes a main focus of therapy when reviewing the assigned homework.
Step 6. Consolidation Phase
The success of a behavioral experiment in large part depends on how effectively the therapist reviews the outcome of the exercise at the following session. Based on information
recorded on the Empirical Hypothesis-Testing Form, the therapist uses a combination of
active listening and probing questions to determine how the exercise was implemented
and the client’s evaluation of the outcome. Rouf et al. (2004) suggest that a number of
issues should be explored including (1) the client’s thoughts and feelings before, during,
and after the experiment; (2) any changes in physical state; (3) evidence that any safety
216
ASSESSMENT AND INTERVENTION STRATEGIES
behaviors or other self-protective measures were utilized; (4) observations about how
other people reacted to the client; (5) significant features of the environment; and (6) the
outcome in terms of noticeable changes in the client’s thoughts and feelings.
When discussing the experiment it is particularly important to evaluate the outcome
in light of the previously stated hypothesis, or predicted outcome. Did the client experience as much anxiety as expected? Was her own response or the responses of others consistent with her prediction? Was the outcome more or less positive than expected? How
similar was the actual outcome to the predicted outcome? If there was a discrepancy,
what does this indicate about the relation between threat appraisals and anxiety? When
reviewing the outcome of a behavioral experiment, the therapist is drawing the client’s
attention to the anxiety-provoking properties of heightened threat and vulnerability
interpretations, and the anxiety-reducing effects of the alternative perspective. The goal
is to reinforce the cognitive conceptualization of anxiety and to promote the idea that
cognitive change is a critical component of anxiety reduction.
The overall purpose of the consolidation phase, then, is to arrive at the significance
or personal meaning of the exercise for clients. Did the behavioral experiment provide a powerful demonstration of the cognitive conceptualization of anxiety? Did they
learn something new about their thinking or way of coping with anxiety that might
be responsible for its reduction? Did the exercise highlight how exaggerated threat and
vulnerability appraisals can intensify subjective anxiety? What can the client take away
from the experiment? This form of questioning will ensure that the behavioral experiment fits within the cognitive case conceptualization developed for the client. It will also
help consolidate any therapy gains that have been made during the sessions. In fact the
main purpose of behavioral experiments that are assigned as homework is to reinforce
or consolidate what has been introduced in the cognitive therapy session by providing the client with personally relevant experiential evidence. It is this critical review of
the experiment’s outcome and its implications that enables empirical hypothesis-testing
exercises to play a significant role in the therapeutic process.
Jodie reported at the following therapy session that she did the behavioral experiment and recorded a description of the experiment and its outcome on the Empirical
Hypothesis-Testing Form. She noted that she arrived to the class at 8:55 and sat in the
middle row. Ten minutes later she made a detailed observation of her peers and then 15
minutes into the lecture she took three to four deep breaths and observed possible reactions. However, she was unable to even slightly shake her body at the 20-minute mark
because of fear someone would notice her odd behavior. In the outcome section of the
form she wrote that only one or two students even glanced at her when she sat down in
class or took the deep breaths. When the therapist reviewed this further, Jodie indicated
she was actually quite surprised that her fellow students paid so little attention to her.
She was also surprised that she actually experienced less anxiety than usual during the
class. The therapist highlighted the discrepancy between the actual outcome (“Students
pay little attention even when Jodie acted in a way that might draw some momentary
attention”) and Jodie’s prediction (“If I do anything like breath differently it will draw
attention to me and I’ll get very anxious”). This experiment was a powerful demonstration for Jodie that thinking others are looking at her makes her more anxious than other
people’s actual momentary glances, and that testing out her anxious thoughts (“people
are looking at me”) with actual evidence (“people take much less notice of me than I
think”) will lead to a reduction in anxious feelings.
Cognitive Interventions for Anxiety
217
Step 7. Findings and Implications Summarized
A final step in empirical hypothesis testing is to summarize the findings and draw out
their implications for developing a new approach to anxiety. This summary statement
can be written on the Empirical Hypothesis-Testing Form and given to the client for
future reference. For Jodie the classroom experiment was summarized in the following
way:
“People often have exaggerated threatening thoughts like ‘everyone in the class is
looking at me and thinking there is something wrong with me.’ These thoughts
are often biased and even untrue yet they cause considerable anxiety. When we put
these thoughts to the test and realize they are not true, our level of anxiety will
decrease substantially. So in the future, when you feel anxious, ask yourself ‘Is my
thinking accurate or am I exaggerating the threat or danger in this situation?’ Test
it out against reality. If there is little evidence to support the thinking, come up with
an alternative view that you can act on.”
Clinician Guideline 6.14
Empirical hypothesis testing is one of the most powerful clinical tools for changing anxious
thoughts, feelings, and behavior. Exercises are designed to test the accuracy of anxious interpretations and reinforce the viability of alterative explanations. Effective behavioral experiments require careful planning and specification that are derived from the cognitive case
formulation. Discussion of the outcome and its implications is an important component of
this therapeutic intervention.
COGNITIVE STRATEGIES IN DEVELOPMENT:
EXPANDING THE CLINICAL ARMAMENTARIUM
The cognitive therapy approach to anxiety disorders is an evolving psychotherapy that
fosters new developments in therapeutic interventions that are derived from empirical
research and clinical experience. There are four new cognitive procedures that have
appeared in the clinical literature that may hold promise in the treatment of anxiety
disorders. Unlike the standard cognitive interventions discussed in the previous section,
these new interventions are still in the development phase and undergoing empirical
investigation. Until more is known about their efficacy and incremental contribution to
cognitive therapy, they should be utilized as auxiliary therapeutic strategies when conducting cognitive therapy for anxiety.
Attentional Training Technique
Wells (2000) introduced the attentional training technique (ATT) as a therapeutic procedure for modifying the perseverative nature of self-referent processing. Highly persistent, repetitive thought is often seen in the anxiety disorders in the form of worry,
obsessions, or anxious rumination. The rationale behind ATT is to teach the anxious
218
ASSESSMENT AND INTERVENTION STRATEGIES
individual how to interrupt repetitive self-attentional processing that contributes to the
persistence of the anxious state. Wells (2000) suggests that ATT may be effective in alleviating emotional distress (e.g., anxiety) by weakening self-focused attention, disrupting
rumination and worry, increasing executive control over attention, and strengthening
metacognitive processing.
According to Wells (2000), ATT consists of auditory attentional exercises in which
clients are taught to selectively attend to neutral noises, rapidly switch their attention
between different sounds, and divide their attention among diverse sounds. The entire
procedure takes 10–15 minutes of therapy time and is practiced in a nonanxious state.
First clients are provided a rationale for ATT. The main point communicated to the
client is that ATT is a procedure for reducing self-focused attention which is known to
intensify anxious thoughts and feelings. The therapist can use specific demonstrations to
illustrate the negative effects of self-focused attention (e.g., have the client intensely focus
on an anxious thought or image and note any changes in mood state). After ensuring that
the rationale has been accepted, the therapist introduces a self-attention rating scale in
which clients use a –3 to +3 bipolar scale to indicate the extent to which their attention
is entirely focused on external stimuli (–3) to their attention is entirely self-focused (+3).
These ratings are administered before and after the ATT practice session to ensure that
the directed attention exercise resulted in a reduction in self-focused attention.
In the actual ATT procedure, the therapist instructs the client to focus on a dot on
the wall. Seating behind the client, the therapist first instructs the client to attend fully
and completely to her voice. Next the client is asked to attend to a tapping sound made
by the therapist. Again the instructions are to shift one’s attention so the client is fully
and completely refocusing on the tapping sound and not letting any other sounds distract him from this task. Then the client is asked to attend to a third sound in the room
such as a ticking clock. This procedure is then repeated for three different sounds in
the near distance (i.e., just outside the room) and three sounds in the far distance (i.e.,
sounds that are outside in the street).
After clients have practiced focusing attention on different sounds, the therapist
calls out the different sounds and they are asked to rapidly shift attention between the
different sounds. This rapid shifting of attention is practiced for a few minutes. Finally,
instructions are given to expand attention by trying to concentrate on all of the sounds
simultaneously and to count the number of sounds heard at the same time. After completing the training procedure, the therapist obtains client feedback. It is emphasized
that intentional direction of attention is difficult but with practice they will become
more proficient. Homework is assigned consisting of 10–15 minutes of ATT practice
twice a day. However, it is important to ensure that clients do not use ATT to avoid their
anxious thoughts or to control anxious symptoms (Wells, 2000).
A variant of ATT that is probably even more applicable to the anxiety disorders
is situational attentional refocusing (SAR). In SAR anxious clients are taught to shift
attention from an internal focus to external information that may disconfirm the
threat-related interpretation. Wells (2000) discusses the use of SAR in conjunction with
exposure in which an individual with social phobia is taught how to shift attention to
external information in the social situation which interrupts the deleterious self-focused
attention that is often seen in social anxiety. For example, when an individual with
social anxiety enters a feared social situation and becomes overly focused on herself
Cognitive Interventions for Anxiety
219
(i.e., self- conscious) and how bad she feels, she is instructed to shift her focus of attention and observe the appearance and facial expressions of other people in the situation.
Note whether these people really are looking at you (Wells, 2000). Although empirical
support for the efficacy of ATT or SAR is still preliminary, findings from a series of
single-case studies are promising (Papageorgiou & Wells, 1998; Wells & Papageorgiou,
1998b; Wells, White, & Carter, 1997).
Metacognitive Intervention
The ability to monitor and regulate our information-processing apparatus is a critical
executive function that is important to human adaptation and survival. We not only
evaluate external stimuli that impinge on our senses, but we also evaluate our own
thoughts and beliefs. Flavell (1979) referred to this capacity to evaluate and regulate
our thinking processes as metacognition, or “thinking about thinking.” Metacognition
is evident as a dynamic cognitive process in which we appraise the thoughts, images,
and impulses that enter the stream of consciousness as well as more enduring beliefs or
knowledge about cognition and its control. Wells (2000) defined metacognition as “any
knowledge or cognitive process that is involved in the appraisal, monitoring or control
of cognition” (p. 6).
An important function of metacognitive processes is the instigation of cognitive
control strategies that could lead to the intensification or shift in internal monitoring
(i.e., conscious awareness) toward or away from a particular thought (Wells, 2000). As
evident from the review in Chapter 3, emotion has a significant biasing effect on information processing. It is conceivable that during anxious states, metacognitive beliefs about
threat are activated and internal monitoring processes become biased toward detection
and elaboration of threat-related thinking. Examples of threat-relevant metacognitive
beliefs include “The more one thinks anxious thoughts, the more likely the feared outcome will happen,” “I’ll become completely overwhelmed with anxiety if I don’t stop
thinking this way,” “If I think it is dangerous, the situation must be dangerous.” In turn
these beliefs could lead to activation of compensatory metacognitive control strategies,
such as efforts to intentionally suppress anxious thoughts, which paradoxically cause
an increase in the salience of the unwanted thoughts and persistence of the negative
emotional state (Wells, 2000, 2009; Wells & Matthews, 2006).
The relevance of a metacognitive conceptualization is clearly evident in OCD and
GAD where individuals engage in obvious appraisals of their unwanted distressing
thoughts (i.e., obsessions, worry) and wage desperate attempts to control the mental
intrusions (see D. A. Clark, 2004; Wells, 2000, 2009, for further discussion). However,
metacognitive beliefs, appraisals, and control strategies are evident in most of the anxiety disorders and so it can be important to intervene at this level when offering cognitive
therapy for anxiety. There are three aspects to cognitive therapy at the metacognitive
level that must be considered.
Metacognitive Assessment
As a first step it is important to identify the primary metacognitive appraisals, beliefs,
and control strategies that characterize the anxious state. Once the main automatic anx-
220
ASSESSMENT AND INTERVENTION STRATEGIES
ious thoughts have been identified, the therapist can probe for metacognitive processes
in the following way.
•฀ “When you have this anxious thought (e.g., ‘I’m going to completely blow this
job interview and never find decent work’), what makes this a significant or a
threatening thought for you?”
•฀ “Are you concerned about any negative consequences from having such
thoughts?”
•฀ “Why do you think you keep having these thoughts?”
•฀ “Is it possible to get control over them? If so, which control strategies work and
which ones don’t work for you?”
Notice that this line of questioning focuses on how the individual appraises the
experience of having anxious thoughts. In the present example, the client may indicate
that he is concerned that having such anxious thoughts before the interview might make
him even more anxious and more likely to perform poorly. A prominent metacognitive
belief might be “Thinking you’ll blow the interview makes it more likely you won’t get
the job” and “It’s critical to get control of this thinking in order to have a good job interview.” Once such metacognitive beliefs and appraisals have been identified, assessment
should focus on the actual mental control strategies that an individual might employ to
shift attention away from the anxious thinking.
Metacognitive Intervention
Having identified the key metacognitive appraisals and beliefs that characterize the anxious state, the cognitive therapist can employ standard cognitive restructuring strategies
to modify this cognitive phenomenon. Strategies such as evidence gathering, cost–benefit
analysis, decatastrophizing, and empirical hypothesis testing can be used to change
metacognitive processes. The difference is not in the interventions but rather in what
is targeted for change. In our previous discussion these cognitive strategies were used
to directly modify the exaggerated threat and vulnerability appraisals that characterize anxious states. In the present discussion these same intervention strategies are used
to modify “thinking about thinking,” that is, the appraisals and beliefs about thought
processes.
To illustrate, an anxious client believes “If I keep thinking I am going to have a car
accident, I’m afraid this way of thinking will actually cause it to happen” (i.e., thought–
action fusion). As a cognitive intervention the client could be asked to examine the
evidence that motor vehicle accidents are caused by anxious thoughts. Inductive reasoning could be used to explore how a thought can lead to a physical catastrophe like a
serious motor vehicle accident. A behavioral exercise could be set up in which the client
observes the effects of such thoughts on her driving behavior or that of other motorists.
A survey could be taken among friends, family, and work associates to determine how
many people thought they would have an accident and then experienced a serious car
accident. These cognitive interventions would focus on modifying the metacognitive
appraisals of significance associated with the “accident premonition” so that the individual begins to interpret such thinking in a more benign fashion such as “the product
of a highly cautious driver.”
Cognitive Interventions for Anxiety
221
Metacognitive Control
An important part of intervention at the metacognitive level is a consideration of the
actual thought control strategies used to deal with unwanted cognition. It is well known
that certain control responses such as the intentional suppression of unwanted thoughts,
rumination, self- critical or punishment responses, neutralization, reassurance seeking,
and thought stopping are ineffective at best and counterproductive at worst (for review
see D. A. Clark, 2004; Wells, 2000, 2009). The cognitive therapist should target any
ineffective control responses used by the client. Cognitive restructuring and empirical
hypothesis-testing exercises may be necessary in order to highlight the deleterious effect
of cherished mental control responses. More adaptive approaches to mental control such
as thought replacement, behavioral distraction, attentional training, or passive acceptance of the thought (e.g., mindfulness) can be introduced in a pragmatic fashion in
order to empirically determine for the client the most effective mental control strategy
to cope with unwanted anxious thoughts.
At this point we have no empirical data to indicate that cognitive therapy that
incorporates a metacognitive perspective is more or less effective than a more standard
cognitive therapy that focuses only on automatic anxious thoughts and beliefs. As will
be seen in a later chapter, the CBT approach to OCD has a strong focus at the metacognitive level and a number of clinical trials have demonstrated its efficacy for OCD. Clinical experience would suggest that evidence of faulty metacognitive appraisals, beliefs,
and control strategies in the persistence of a client’s anxiety disorder would warrant a
greater focus on these processes in therapy.
Imaginal Reprocessing and Expressive Writing
Although memories of past traumatic experiences are a prominent diagnostic feature of
PTSD (DSM-IV-TR; American Psychiatric Association [APA], 2000), recollections of
highly anxious experiences can play a key role in the persistence of any anxiety disorder.
In fact threatening visual images of past experiences or anticipated possibilities in the
future are common in all the anxiety disorders (Beck et al., 1985, 2005). These anxious
fantasies or past recollections are often a biased and distorted representation of reality that can fuel an anxious state. For example, in panic disorder an individual might
imagine a horrible death via suffocation, a person with social anxiety might remember
a past experience of trying to speak up in a group of unfamiliar people, someone with
OCD might recall a vivid memory of touching something quite disgusting and feeling
a profound sense of contamination, or the individual with GAD might imagine her life
after experiencing a financial disaster. In each of these cases the therapist should include
imagery or memory modification as a therapeutic goal for treatment.
Modification of anxious memories or imagery begins with clients providing a full
and detailed account of their memory or anxious fantasy. The therapist should elicit
all relevant automatic thoughts, beliefs, and appraisals that constitute the biased threat
interpretation of the memory or anticipated event. Descriptions of reliving approaches
to traumatic memories in CBT for PTSD suggest a number of methods for enhancing
clients’ exposure to traumatic memories or anxious images and dealing with elevated
anxiety levels (e.g., Foa & Rothbaum, 1998; Ehlers & Clark, 2000; Shipherd, Street,
& Resick, 2006; Taylor, 2006). Extensive discussion and therapeutic questioning is an
222
ASSESSMENT AND INTERVENTION STRATEGIES
obvious initial step in exposure. This is followed by asking clients to write out a narrative of the traumatic memory or imagined catastrophe (for further discussion, see
Chapter 12 on PTSD). This narrative should be as detailed as possible so it can be used
as the basis of repeated exposure to the traumatic memory (i.e., reliving the experience).
Standard cognitive restructuring strategies are employed to modify faulty appraisals and beliefs associated with the memory or imagined catastrophe (Ehlers & Clark,
2000). The goal is to arrive at an alternative perspective toward the memory or anxious
fantasy that is more adaptive and less anxiety-provoking. In addition, efforts should be
made to construct a more balanced memory of the traumatic experience itself that is a
closer approximation to reality. For individuals who are troubled by images of anticipated catastrophe, again a more realistic scenario can be developed. The client can be
encouraged to practice replacing the maladaptive memory or fantasy with the more
adaptive alternative. Behavioral exercises can be assigned that would strengthen the
alternative memory or fantasy and weaken the traumatic recollection or anxious imagery. Given the extensive use of cognitive restructuring and construction of an alternative
perspective, this form of imaginal intervention is better described as a “reprocessing
intervention” (i.e., a reprocessing of the memory or anxious fantasy) rather than simply
repeated exposure to an internal fear stimulus.
The contribution of memory or imagery reprocessing to the effectiveness of cognitive treatment for the anxiety disorders is unknown. Research that has focused specifically on the active ingredients of CBT for PTSD indicates that imaginal and situational
exposure are critical components of the treatment’s effectiveness (see review by Taylor,
2006). Moreover, Pennebaker (1993) found that thinking and talking about a traumatic
event immediately after its occurrence is an important phase in the natural adaptation
to traumatic events. More recently, Pennebaker and colleagues demonstrated that a relatively brief intervention in which individuals write on their deepest thoughts and feelings about an emotional upheaval produces positive emotional, behavioral, and healthrelated benefits including reductions in depressive symptoms for individuals who tend
to suppress their thoughts (e.g., Gortner, Rude, & Pennebaker, 2006; see Pennebaker,
1997; Smyth, 1998). These findings, then, suggest that modification of highly distressing
memories of past experiences or fantasies of future catastrophes is an important target
for cognitive intervention when this phenomena plays a critical role in the maintenance
of an individual’s anxiety state.
Mindfulness, Acceptance, and Commitment
Segal, Williams, and Teasdale (2002) describe an eight-session group intervention for
individuals who recovered from major depression aimed at reducing depressive relapse
through training in mindfulness approaches that help individuals “decenter” from their
negative thinking. Called mindfulness-based cognitive therapy (MBCT), the intent is
to teach individuals a different way to become aware of and relate to their negative
thinking. Rather than become engaged with their negative cognitions in an evaluative
manner, individuals are taught to “decenter” from their thoughts, feelings, and bodily
sensations. That is, negative thoughts are to be observed and described but not evaluated (Segal, Teasdale, & Williams, 2005). Group participants are taught to focus their
awareness on their experience in the moment in a nonjudgmental manner. Eight-week
Cognitive Interventions for Anxiety
223
2-hour group sessions guide participants in exercises that increase moment-by-moment
nonjudgmental awareness of bodily sensations, thoughts, and feelings. Daily homework
in awareness exercises is a critical component of the treatment. The rationale behind
mindfulness approaches is that a nonjudgmental “decentered” approach will counter the
automatic patterns of cognitive-affective processing that can lead to depressive relapse
(Segal et al., 2005).
Although clinical trials on the efficacy of MBCT are at a preliminary stage, there is
evidence that the intervention can significantly reduce depressive relapse rates in those
with three or more previous episodes of major depression compared with a treatment
as usual condition (Ma & Teasdale, 2004; Teasdale et al., 2000). Furthermore, MCBT
was most effective in preventing relapse/recurrence of episodes that were unrelated to
negative life experiences. Since MBCT is an adaptation of Jon Kabat-Zinn’s mindfulness
meditation that has been used extensively at the University of Massachusetts for reduction of stress, pain, and anxiety, it has obvious relevance for treatment of anxiety disorders (see Germer, 2005; Kabat-Zinn, 1990, 2005; Kabat-Zinn et al., 1992). In a pilot
study 14 patients with panic disorder and eight with GAD received an 8-week group
meditation-based stress reduction and relaxation program (Kabat-Zinn et al., 1992).
Twenty patients showed significant reduction in BAI and Hamilton Anxiety scores at
posttreatment and a significant decrease in panic attacks. Although these preliminary
findings are encouraging, full randomized controlled clinical trials will be needed before
the full implication of mindfulness interventions for countering anxious cognition and
reducing anxiety states is known.
Acceptance and commitment therapy (ACT), introduced by Dr. Steven Hayes, is
a psychotherapeutic perspective linked to post- Skinnerian radical behaviorism that
focuses on the context and function of psychological phenomena (i.e., cognition) rather
than on its form and content (Hayes, 2004). ACT is based on an underlying philosophy
of functional contextualism in which the function of phenomenon (e.g., a worrisome
thought) is understood in terms of the whole organism interacting within a historical
and situational context (Hayes, 2004; Hayes, Strosahl, & Wilson, 1999). The goal of
functional contextualism is the prediction and influence of events that lead to psychological flexibility, that is, the ability to change or persist with functional behaviors that
serve valued ends (Hayes, 2004). The following is a brief description of the six core
therapeutic processes in ACT (for more detailed discussion, see Hayes, Follette, & Linehan, 2004; Hayes & Strosahl, 2004; Hayes, Strosahl, Buting, Twohig, & Wilson, 2004;
Hayes et al., 1999).
•฀ Acceptance —an openness to experience thoughts and feelings with nonjudgmental awareness; to embrace thoughts and feelings as they are rather than as events that
must be controlled or changed. Clients learn through various experiential and mindfulness exercises to psychologically accept even their most intense thoughts, feelings, and
bodily sensations.
•฀ Cognitive defusion —refers to the process of objectifying thoughts so that thoughts
are viewed as merely thoughts and no longer fused with the self or personal experience.
A variety of techniques can be used to help clients defuse or separate themselves from
the literal meaning of thoughts such as having clients repeatedly verbalize a difficult
thought until it is merely heard without meaning or evaluation, or watch thoughts as
external objects without use or involvement (Luoma & Hayes, 2003).
224
ASSESSMENT AND INTERVENTION STRATEGIES
•฀ Self as context—ACT focuses on helping clients release their attachment to an
unhealthy conceptualized self and embrace a transcendent sense of self through a variety of mindfulness/meditation, experiential exercises, and metaphors (Hayes, Follette,
et al., 2004).
•฀ Being present—this refers to the promotion of an active, open, effective, and
nonjudgmental awareness or contact with the present moment rather than fusion and
avoidance which interfere with “being present in the moment.”
•฀ Values—clients are encouraged to select and clarify their fundamental life values
which can be described as “chosen qualities of purposive action” (Hayes, Follette, et al.,
2004). For example, clients can be asked what they would like to see written on their
tombstone.
•฀ Committed action —this involves choosing specific goals and then taking responsibility for behavioral changes, adapting and persisting with behavioral patterns that
will lead to desired goals. Various intervention strategies such as psychoeducation, problem solving, behavioral homework, skills training, and exposure can be used to achieve
committed action (Hayes, Follette, et al., 2004).
There are fundamental differences between ACT and cognitive therapy in their
view of cognition. In cognitive therapy the term cognition refers to a thought process,
whereas ACT considers it private behavior and so focuses on changing its function
rather than its content (Hofmann & Asmundson, 2008). Furthermore, Hofmann and
Asmundson (2008) note that the two approaches differ in their emotion regulation strategy, with cognitive therapy emphasizing change in the antecedents of emotion and ACT
focusing on experiential avoidance or the response side of emotion regulation. This leads
to fundamental differences in therapeutic approach, with ACT using mindfulness and
other strategies to teach a nonevalutive, nonjudgmental approach to negative thoughts
that encourage their acceptance and integration into a wide variety of actions (Luoma
& Hayes, 2003). Of course, cognitive therapy emphasizes the evaluation and correction
of negative thought content through cognitive and behavioral intervention strategies.
According to ACT, the main problem in the anxiety disorders is experiential avoidance, that is, an unwillingness to experience anxiety including its attendant thoughts,
feelings, behaviors, and bodily sensations (Orsillo, Roemer, Lerner, & Tull, 2004). As
a result, anxious individuals struggle against their anxiety, relying on ineffective and
futile external and internal control strategies as well as escape and avoidance to alleviate
the unacceptable anxiety. The goal of ACT is the reduction of experiential avoidance,
which prevents the attainment of valued goals by teaching the anxious person experiential acceptance defined as “a willingness to experience internal events, such as thoughts,
feelings, memories, and physiological reactions, in order to participate in experiences
that are deemed important and meaningful (Orsillo et al., 2004, p. 76).
Orsillo and colleagues describe a 16-session individual ACT/mindfulness intervention for GAD that promotes experiential acceptance of anxiety through training in mindfulness, acceptance, cognitive defusion, meditation, relaxation, and self-monitoring. In
addition an emphasis is placed on defining life values that have been impeded by experiential avoidance and commiting to behavioral changes that focus on valued activities
so that the individual is behaving intentionally rather than reactively. In an open trial
Roemer and Orsillo (2007) reported that ACT led to significant reductions on measures
of GAD severity, worry, general anxiety, and stress symptoms that were maintained at
Cognitive Interventions for Anxiety
225
3-month follow-up. Twohig, Hayes, and Masuda (2006) utilized a multiple-baseline,
across-participant research design involving eight weekly 1-hour sessions of ACT to
demonstrate treatment effectiveness in four individuals with OCD. However, a recent
meta-analysis of various “third wave” therapies, including ACT, concluded that their
mean effect sizes were only moderate, the outcome studies lacked the methodological
rigor seen in CBT, and so they fail to meet criteria for empirically supported treatments
(Öst, 2008). It may be that a greater focus on training the anxious person to adopt a
nonevaluative, benign acceptance and distancing perspective on anxious thinking has
clinical utility in the treatment of the anxiety disorders, but this conclusion must await
the results of more rigorous treatment outcome research.
Clinician Guideline 6.15
Attentional training may be used to interrupt heightened self-focused attention, whereas
cognitive restructuring strategies can be redirected toward modification of faulty metacognitive processes and thought control strategies. Imaginal reprocessing and expressive writing may be helpful in modifying memories of past traumatic experiences or imagined future
catastrophes, whereas mindfulness and cognitive diffusion derived from ACT may be used
to teach clients a more detached, nonevaluative approach to anxious cognitions. Although
promising, these approaches lack the strong clinical and empirical base of standard cognitive interventions for anxiety.
SUMMARY AND CONCLUSION
Modification of the exaggerated appraisals of threat, vulnerability, and safety seeking is
the primary objective of cognitive therapy for anxiety disorders. This chapter presented
the main cognitive strategies that comprise cognitive treatment protocols developed for
the specific anxiety disorders. These strategies are entirely consistent with the cognitive
model of anxiety (see Figure 2.1) and they target the aberrant cognitions identified in
the case formulation.
The goal of any cognitive intervention is deactivation of the hypervalent threat
schemas and heightened activation of more adaptive and realistic beliefs about threat
and perceived ability to cope with one’s anxious concerns. This is achieved by shifting
the client’s focus away from threat content and onto the faulty appraisals and beliefs
that are the basis of the anxious state. Exaggerated appraisals of the probability and
seriousness of threat are targeted as well as the heightened evaluations of personal vulnerability and need to seek safety. Cognitive interventions also seek to increase personal
self-efficacy for dealing with anxiety by normalizing the fear response and fostering a
more adaptive perspective on the balance between risk and safety.
A detailed description was provided on how to implement the main cognitive strategies that define this treatment approach to anxiety. Educating the client into the cognitive model of anxiety is an important first step in establishing therapeutic collaboration
and compliance with treatment. Teaching self-monitoring skills in the identification of
automatic anxious thoughts and appraisals, though critical to the success of cognitive
226
ASSESSMENT AND INTERVENTION STRATEGIES
therapy, can be especially difficult given the heightened emotional state and situational
specificity of anxiety. However, once an awareness of exaggerated threat appraisals has
been established, cognitive restructuring strategies such as evidence gathering, cost–
benefit analysis, and decatastrophizing can be utilized to challenge anxious schemas.
Teaching the anxious individual to become much more aware of cognitive errors
and faulty inductive reasoning during periods of intense anxiety helps foster a more
critical attitude toward one’s anxious thinking style. Formulating alternative perspectives on anxious situations and concerns that bear a closer approximation to reality
offers a counterpoint to the exaggerated threat and vulnerability that characterizes
anxiety. However, the most powerful tool in the cognitive therapist’s armamentarium is
the behavioral experiment or empirical hypothesis-testing exercise. Behavioral exercises
provide clients with experiential data that refute threat and vulnerability schemas and
support an alternative, adaptive perspective. A Quick Reference Summary is provided
in Appendix 6.6 to remind the clinician of various cognitive strategies available for
therapeutic intervention.
APPENDIX 6.1
Selected List of Self-Help Treatment Manuals That Can Be Assigned When Educating
a Client into the Cognitive Model and Treatment of Anxiety
1. Abramowitz, J. S. (2009). Getting over OCD: A 10-step workbook for taking back your life.
New York: Guilford Press.
2. Antony, M. M., & McCabe, R. E. (2004). 10 simple solutions to panic: How to overcome panic
attacks, calm physical symptoms and reclaim your life. Oakland, CA: New Harbinger.
3. Antony, M. M., & Norton, P. J. (2008). The anti-anxiety workbook: Proven strategies to
overcome worry, phobias, panic and obsessions. New York: Guilford Press.
4. Antony, M. M., & Swinson, R. P. (2000b). The shyness and social anxiety workbook: Proven
techniques for overcoming your fears. Oakland, CA: New Harbinger.
5. Barlow, D. H., & Craske, M. G. (2007). Mastery of your anxiety and panic: Workbook (4th ed.).
Oxford, UK: Oxford University Press.
6. Butler, G., & Hope, T. (2007). Managing your mind: The mental fitness guide. Oxford, UK:
Oxford University Press.
7. Clark, D. A., & Beck, A. T. (2010). Defeat fear and anxiety: A cognitive therapy workbook.
Manuscript in preparation. Department of Psychology, University of New Brunswick, Canada.
8. Craske, M. G., & Barlow, D. H. (2006). Mastery of your anxiety and worry: Workbook (2nd ed.).
Oxford, UK: Oxford University Press.
9. Hope, D. A., Heimberg, R. G., Juster, H. R., & Turk, C. L. (2000). Managing social anxiety: A
cognitive-behavioral therapy approach. Client workbook. Oxford, UK: Oxford University Press.
10. Hope, D. A., Heimberg, R. G., & Turk, C. L. (2006). Managing social anxiety: A cognitivebehavioral therapy approach. Oxford, UK: Oxford University Press.
11. Kabat-Zinn, J. (1990). Full catastrophe living: Using the wisdom of your body and mind to face
stress, pain, and illness. New York: Bantam Dell.
12. Leahy, R. L. (2005). The worry cure: Seven steps to stop worry from stopping you. New York:
Harmony Books.
13. Leahy, R. L. (2009). Anxiety free: Unravel your fears before they unravel you. Carlsbad, CA:
Hay House.
14. Purdon, C., & Clark, D. A. (2005). Overcoming obsessive thoughts: How to gain control of your
OCD. Oakland, CA: New Harbinger.
15. Rygh, J. L., & Sanderson, W. C. (2004). Treating generalized anxiety disorder: Evidenced-based
strategies, tools, and techniques. New York: Guilford Press.
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
227
APPENDIX 6.2
Testing Anxious Appraisals: Looking for Evidence
Name:
Date:
1. Briefly state the anxious thought or appraisal:
2. State how likely this outcome feels to you when you are most anxious from 0% (won’t happen ) to 100% (certain):
%
3. State how serious the outcome feels to you when you’re anxious from 0 (not serious) to 100 (a catastrophe):
%
Evidence for the Anxious Thought or Appraisal
Evidence against the Anxious Thought or Appraisal
228
1.
1.
2.
2.
3.
3.
4.
4.
5.
5.
6.
6.
* Use additional pages to list evidence for and against.
4. State how likely this outcome appears after looking at the evidence from 0% (won’t happen ) to 100% (certain):
%
5. State how serious the outcome appears after looking at the evidence from 0 (not serious) to 100 (a catastrophe):
%
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright 2010 by The Guilford Press. Permission to photocopy
this appendix is granted to purchasers of this book for personal use only (see copyright page for details).
APPENDIX 6.3
Cost–Benefit Form
Name:
Date:
1. Briefly state the anxious thought, belief, or appraisal:
Immediate and Long-Term Advantages
Immediate and Long-Term Disadvantages
1.
1.
2.
2.
3.
3.
4.
4.
5.
5.
6.
6.
*Circle the costs and benefits that are most important to you.
2. Briefly state an alternative perspective:
Immediate and Long-Term Advantages
Immediate and Long-Term Disadvantages
1.
1.
2.
2.
3.
3.
4.
4.
5.
5.
6.
6.
*Circle the costs and benefits that are most important to you.
tFrom Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
229
APPENDIX 6.4
Alternative Interpretations Form
Name:
Date:
1. Briefly state the most dreaded outcome (worst-case scenario) associated with your anxiety:
2. Briefly state the most desirable outcome (best possible scenario) associated with your anxiety:
3. Briefly state the most realistic (probable) outcome associated with your anxiety:
Evidence for the Dreaded Outcome
(catastrophic view)
Evidence for the Most Desired Outcome
(most desired goal)
Evidence for the Most Probable Outcome
(alternative view)
230
1.
1.
1.
2.
2.
2.
3.
3.
3.
4.
4.
4.
5.
5.
5.
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright 2010 by The Guilford Press. Permission to photocopy
this appendix is granted to purchasers of this book for personal use only (see copyright page for details).
APPENDIX 6.5
Empirical Hypothesis-Testing Form
Name:
Date:
1. State the threat interpretation associated with your anxiety:
2. State the alternative interpretation proposed in therapy:
3. State the hypothesis (predicted outcome) for this exercise:
Description of the Exercise
Record How Exercise Was Conducted
Describe Outcome of the Exercise
231
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright 2010 by The Guilford Press. Permission to photocopy
this appendix is granted to purchasers of this book for personal use only (see copyright page for details).
APPENDIX 6.6
Chapter 6 Quick Reference Summary: Cognitive Interventions
I. Education Phase (sessions 1–2)
Define anxiety and fear; fear adaptive; cognitive basis of anxiety (handout Fig. 6.1) and use client
examples from assessment; negative consequences of anxiety; role of avoidance and safety seeking (use
client examples); establish treatment goals and CT rationale (turn off, deactivate the “anxiety program”).
II. Identifying the First Apprehensive Thoughts (sessions 2–3)
1. Review client’s “Situational Analysis Form” (Appendix 5.2); probe for immediate, automatic, first
anxious thought. If needed use illustration of “walk alone and hear a noise.”
2. Emphasize the exaggerated probability and severity of threat appraisals in first anxious thinking.
3. Assign “Apprehensive Thoughts Self-Monitoring Form” (Appendix 5.4) as homework.
4. Emphasize importance of homework (see explanation in Chapter 6, pages 199–200) and therapeutic
benefits of understanding one’s anxiety.
III. Standard Cognitive Interventions (sessions 3 to end)
1. Evidence Gathering—first use “Testing Anxious Appraisals: Looking for Evidence” form (Appendix
6.2) in session; use client anxiety episode from past week or from “Situational Analysis Form.” Assign
“Testing Anxious Appraisals” form as homework.
2. Cost–Benefit Analysis—first use “Cost–Benefit Form” (Appendix 6.3) in session; list advantages/
disadvantages of “threat perspective” first and then repeat for “alternative perspective.”
3. Decatastophizing—explore with client his worst outcome; go through preparation, catastrophe
description, and problem-solving stage; have client imagine the worst possible outcome or write down
its description.
4. Identify Thinking Errors—provide client handout of “Common Errors and Biases in Anxiety”
(Appendix 5.6) and go over recent anxious thinking for possible errors; assign “Identifying Anxious
Thinking Errors” as homework.
5. Generating Alternative Explanation—first work on generating alternative thinking to recent anxious
episode; use “Alternative Interpretations Form” (Appendix 6.4); work on evidence for worst outcome,
then most desired outcome, and finally most realistic outcome. Assign as homework if another
anxiety concern is evident.
6. Empirical Hypothesis Testing (homework assignment)—provide rationale; specific statement of threat
appraisal and its competing alternative; plan the experiment (write down instructions); client uses
“Empirical Hypothesis-Testing Form” (Appendix 6.5) to record actual experiment (write down threat
interpretation, alternative, and expected outcome when setting up experiment); explore outcome of
experiment in following sessions (consolidation phase); write out a summary of conclusions about the
experiment for client.
IV. Alternative Cognitive Interventions (latter part of therapy)
1. Attentional Training Technique (ATT)—counters self-focused attention, rumination, and worry;
trained attention to three neutral sounds in office, then three sounds outside office, then three sounds
in distance, use ATT rating scale after each; therapist calls out different sounds to practice alternating
attention; homework assignment is 10–15 minutes of ATT practice twice daily.
2. Metacognitive Intervention—assess whether client engaged in faulty appraisals and beliefs about
her thoughts; use standard cognitive interventions to challenge metacognitive appraisals and beliefs;
encourage cessation of any counterproductive thought control strategies; allow anxious thinking to
“fade naturally.”
(cont.)
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
232
APPENDIX 6.6 (page 2 of 2)
3. Imaginal Reprocessing and Expressive Writing—have client generate script or imagery of traumatic
or troubling imagery or memory; develop an alternative, more adaptive version and repeatedly expose;
client instructed to write out a detailed description of thoughts and feelings associated with past
troubling memory or imagery in form of expressive writing.
4. Mindfulness and Acceptance—utilize self-monitoring and mindfulness exercises to train clients in a
nonjudgmental, observational, and objectifying acceptance of anxious thoughts, feelings, and bodily
sensations in order to reduce experiential avoidance of anxiety.
233
Chapter 7
Behavioral Interventions
A Cognitive Perspective
Courage is resistance to fear, mastery of fear—
not absence of fear.
—M ARK TWAIN (19th- century American author
and humorist, 1835–1910)
Maria had struggled with severe and incapacitating generalized social phobia
since the age of 13. After 18 years of poor response to various medication regimens, hospitalization, and false starts with different psychotherapists, Maria’s
anxiety disorder had worsened to the point where she was practically housebound, unable to work or socialize in a meaningful way. Although there was
evidence of a past comorbid bipolar I disorder, it was the social anxiety that
was the primary diagnosis at the time of assessment. She did not meet diagnostic criteria for current mania or depression, so the intervention focused on her
social anxiety symptoms and associated panic attacks.
Maria had an intense fear of negative evaluation from others, especially
familiar people. She was concerned that others would stare at her and conclude
that she was “nothing” because of her poor physical appearance or because
she had achieved so little with her life. She became preoccupied with her physical appearance and attire, afraid that others would think she was wearing a
“horrible outfit” and so conclude that she was unable to take care of herself.
She developed an intense fear of meeting people from her past who she feared
would remember her inappropriate behavior during past manic episodes and
this would contribute to their harsh judgment of her. When in public settings,
Maria would frequently experience panic attacks that included chest pain,
numbness, smothering sensations, dizziness, and heart palpitations.
In an effort to reduce her heightened state of anxiety, Maria developed a
number of behavioral coping strategies. She avoided all social gatherings and
most public places, leaving her practically housebound. She spent hours getting
ready in the morning in order to look “just perfect” and would compulsively
234
Behavioral Interventions
235
check her appearance in the mirror and seek reassurance from family members
on whether she looked neat and tidy. She was convinced that if she looked perfect, people would think she was competent and this would make her feel more
confident and less anxious. When she started to feel panicky around others,
Maria would engage in an exaggerated form of controlled breathing that was so
extreme that others could not help but notice an unusual breathing pattern that
bordered on hyperventilation. She was also so internally focused on her anxiety that she had difficulty maintaining a conversation. She engaged in extensive
postevent processing in which she would spend considerable time ruminating
about her performance in a social situation. In the end she performed poorly
in social encounters because of her heightened anxiety, panic, and preoccupation. This daily battle against anxiety and perceived social incompetence left
Maria feeling hopeless and pessimistic, drained of all self- confidence and sense
of self-worth.
This case provides a good illustration of the importance of behavior change in
alleviating anxiety disorders. Avoidance, compulsive checking, reassurance seeking,
hyperventilation, and social skills deficits were just some of the maladaptive behavioral responses that actually contributed to the persistence of Maria’s social anxiety. It
was clear from the case formulation that an effective cognitive intervention must focus
on behavioral change. Graded exposure, behavioral experimentation, and social skills
training through use of videotaped feedback and role plays would be critical therapeutic
ingredients in her treatment plan.
In this chapter we discuss the role of behavioral interventions in cognitive therapy
for anxiety disorders. We begin by considering the importance of behavioral strategies in cognitive therapy of anxiety and how these interventions are restructured to
facilitate change in anxious thoughts and beliefs. Attention is then turned to exposure
as the single most effective intervention for therapeutic change across the anxiety disorders. General guidelines and procedures for implementing exposure-based treatment
are considered along with its three main areas of focus: situations, imagery, and physical sensations. We then consider the importance of response prevention in eliminating
maladaptive safety seeking and other forms of ineffective coping responses. Relaxation
and breathing retraining are discussed as possible supplementary elements of cognitive
therapy for anxiety.
IMPORTANCE OF BEHAVIORAL INTERVENTION
Given the prominence of escape and avoidance responses in most forms of pathological anxiety, it is not surprising that behavioral change is a critical aspect of cognitive
therapy for anxiety. Beck et al. (1985, 2005) devoted an entire chapter to behavioral
strategies and behavioral change is emphasized in CBT protocols for specific anxiety
disorders like panic (D. M. Clark, 1997; Craske & Barlow, 2001), social phobia (D. M.
Clark, 2001; Rapee & Heimberg, 1997), OCD (D. A. Clark, 2004; Rachman, 2006;
Salkovskis, 1999; Salkovskis & Wahl, 2003), and PTSD (Ehlers & Clark, 2000; Taylor, 2006). In addition, empirical research indicates that behavioral interventions like
exposure and response prevention have their own direct significant effects on reducing
236
ASSESSMENT AND INTERVENTION STRATEGIES
anxiety (Abramowitz, Franklin, & Foa, 2002; Fava, Zielezny, Savron, & Grandi, 1995;
Feske & Chambless, 1995; Riggs, Cahill, & Foa, 2006). Thus behavioral intervention
strategies are a central therapeutic ingredient of cognitive therapy for anxiety.
Cognitive Perspective on Behavioral Interventions
In cognitive therapy behavioral strategies are employed as interventions for modifying
faulty threat and safety appraisals and beliefs. Thus the cognitive therapist conceptualizes behavioral-oriented assignments quite differently from a strictly behavioral perspective. Instead of viewing behavioral interventions in terms of strengthening inhibition or
habituation of an anxiety response, cognitive therapy views the interventions in terms
of its effect on changing threat-related cognition, which in turn will lead to a reduction
in anxious symptoms. This cognitive reconceptualization of behavioral treatment has
several practical implications for how behavioral interventions are implemented in the
following steps. (See the section on empirical hypothesis testing in the previous chapter
for a discussion of issues relevant to the use of behavioral interventions in cognitive
therapy.)
Rationale
As with any therapeutic intervention, the client should be provided a rationale for the
behavioral assignment that is based on the cognitive model of anxiety presented during
the psychoeducational phase of treatment (see Figure 6.1). There are two essential ideas
about behavioral interventions that should be communicated to clients. First, the cognitive therapist explains that one of the most effective ways to change anxious thinking
is through direct experience with anxiety-provoking situations. In our case example it
was explained to Maria that the experience she gained from exposure to actual social
situations was the most potent way to learn whether other people were evaluating her
as harshly as she imagined.
Second, a cognitive rationale for behavioral interventions should include a discussion of potentially maladaptive behavioral coping strategies. It is explained that modification of these coping strategies is an essential component of cognitive therapy. Another
reason for behavioral interventions, then, is the modification of dysfunctional coping
responses and the acquisition of more effective responses that will lead to a reduction
in anxiety.
Identify Target Thought/Belief
The cognitive therapist always introduces a behavioral intervention as a means for
achieving cognitive change. Thus a specific anxious thought, appraisal, or belief is identified as the primary target for the behavioral intervention. In order for the behavioral
exercise to be effective, the client must be clear on the anxious thought or belief that
is under evaluation by the intervention. For Maria three core beliefs were particularly
critical in her cognitive therapy: “If I happen to meet familiar people, they will consider
that I have little worth or value, that I’m a real failure in life,” “Familiar people will
view me as emotionally unstable because they will remember my ‘crazy’ behavior when
Behavioral Interventions
237
I was manic,” and “If my physical appearance is perfect, people will think I am more
competent and in control.”
Behavioral Prescription
The client is always provided specific information on how to perform a behavioral exercise, something analogous to a behavioral prescription. A schedule indicating when to
do the exercise, where, and for how long should be worked out. It should be clearly
spelled out whether there are restrictions on the use of safety cues (e.g., a person with
agoraphobia can take a trusted friend to the shopping mall but must spend 30 minutes
in the mall alone). Moreover, the therapist should discuss with the client what coping responses are considered healthy when performing the behavioral task and what
responses would undermine the success of the intervention (see section on planning
behavioral experiments in previous chapter).
Self-Monitoring
Clients should record the outcome of any behavioral exercise performed as a homework
assignment. Specific self-monitoring forms should be used such as the assessment or
thought record forms reproduced in the appendices of Chapters 5 and 6 or the behavioral forms that can be found later in this chapter. Although some clients insist on
keeping less formal, more open-ended records of their homework, it is important that
sufficient information is recorded to allow an evaluation of the behavioral assignment
(see previous chapter on recording in behavioral experiments).
Evaluation
The postintervention follow-up is perhaps the most critical component of the behavioral exercise in cognitive therapy. The therapist should review in detail the information
recorded on the self-monitoring form. It is critical to highlight how the client’s experience with the behavioral intervention disconfirmed the anxious appraisal and supported
an alternative interpretation. This could even be written down on a “coping card” that
clients use to counter their anxious thoughts in subsequent anxious episodes.
In our case example Maria was asked to accompany a friend to a café and sit with
her for at least 20 minutes while they had a drink and chatted about their daily lives.
Maria was asked to self-monitor her anxiety level throughout the behavioral assignment, taking particular notice of her automatic thoughts and any social cues that she
picked up from those around her. She made two important observations. First, her
anxiety escalated even further as she became more and more preoccupied with her internal anxious state and worried that others noticed that she looked uncomfortable. And
second, there was no objective evidence that anyone even noticed her in the café. No
one was looking at her or showed the least interest in her presence. Thus the behavioral
experiment disconfirmed her maladaptive belief that her anxiety was due to others looking at her, of being the “center of their attention,” and supported the alternative explanation that her anxiety was due to heightened self-focused attention on her internal
state. Based on the results of this assignment, therapy then focused on various cognitive
238
ASSESSMENT AND INTERVENTION STRATEGIES
strategies to counter the deleterious effects of heightened self-focused attention when in
social situations.
Clinician Guideline 7.1
Behavioral interventions are a critical therapeutic ingredient of cognitive therapy of anxiety. These interventions are used to directly test the dysfunctional thoughts and beliefs
that maintain anxiety. Behavioral interventions are introduced early in treatment and used
throughout therapy in a highly structured and organized fashion as within-session demonstrations and between-session homework assignments.
EXPOSURE INTERVENTIONS
Exposure involves systematic, repeated, and prolonged presentation of objects, situations, or stimuli (either internal or external) that are avoided because of their anxietyprovoking properties. The effectiveness of in vivo exposure has been clearly demonstrated
for panic disorder, with situational exposure essential when agoraphobic avoidance
is present (van Balkom, Nauta, & Bakker, 1995; Gould, Otto, & Pollack, 1995). In
addition, exposure is an effective intervention strategy for OCD (see Foa, Franklin, &
Kozak, 1998; Foa & Kozak, 1996), social phobia (Heimberg & Juster, 1995), and PTSD
(Foa & Rothbaum, 1998; Riggs et al., 2006). Exposure, then, is one of the most powerful therapeutic tools available to the therapist for the reduction of fear and anxiety.
Exposure procedures are effective because they modify fear memory structures.
Foa and Kozak (1986) contend that exposure must present fear-relevant information
that fully activates the fear memory structure. Exposure information that is sufficiently
incompatible with meaning and response elements of the fear structure will lead to a
decrease in fear and anxiety, whereas information compatible with the fear structure
will have the opposite effect. Two important therapeutic implications can be drawn
from this analysis.
1. Effective exposure must activate fear schemas (i.e., memory structures). In other
words, individuals must be moderately anxious during the exposure exercise in
order to attain therapeutic threshold.
2. Effective exposure must present disconfirming information. The success of an
exposure experience will depend on whether the individual is fully attentive to
and processes incompatible information that disconfirms exaggerated threat and
vulnerability elements of the fear schema.
In addition to a solid theoretical and empirical basis for exposure, these procedures
serve multiple functions within cognitive therapy for anxiety. Table 7.1 presents a summary of the reasons for using exposure in cognitive therapy of anxiety.
Three types of exposure interventions can be utilized in fear reduction: in vivo
or situational, imaginal, and internal exposure. Situational exposure involves contact
with physical objects or actual situations that are avoided in the external environment,
Behavioral Interventions
239
TABLE 7.1. Purpose of Exposure in Cognitive Therapy of Anxiety
Reasons for including exposure procedures in cognitive therapy
•฀ To provide assessment information on the anxiety response in avoided situations
•฀ To provide corrective information that disconfirms perceived threat and vulnerability
•฀ To test catastrophic beliefs through behavioral experimentation
•฀ To confirm alternative, more adaptive appraisals and beliefs
•฀ To reinforce adaptive coping strategies and challenge the utility of maladaptive responses
•฀ To weaken reliance on safety-seeking cues and behavior
•฀ To provide new learning experiences about fear and anxiety
•฀ To reduce or eliminate escape and avoidance behavior
whereas internal self-focused procedures involve exposure to feared physical sensations
(Antony & Swinson, 2000a). Imaginal exposure involves presentation of symbolic fear
stimuli. Later we will discuss the implementation of each of these exposure procedures,
but first we consider a number of issues that must be addressed when undertaking an
exposure-based intervention.
General Guidelines for Exposure Procedures
Probably no other psychotherapeutic intervention has been misjudged more often
than exposure-based treatment. The intervention appears deceptively simple and yet
most therapists can attest to the difficulty of its implementation. Ensuring that clients
receive sufficient “dosage” to be therapeutically effective is a challenge in its own right.
Many individuals give up after one or two exposure attempts so their experiences only
heighten rather than reduce anxiety. The following issues must be taken into account
when planning an exposure intervention. (For an expanded discussion of guidelines for
implementing exposure procedures, see Antony & Swinson [2000a]; Craske & Barlow
[2001]; Foa & Rothbaum [1998]; Kozak & Foa [1997]; Steketee [1993]; and Taylor
[2000, 2006].)
Rationale and Planning
The cognitive therapist explains exposure procedures as effective interventions that
provide direct experience with information that disconfirms anxious appraisals and
beliefs. It is emphasized that learning from experience has a much more powerful effect
on changing emotion-based thinking than logical persuasion. However, some clients
might express skepticism about the therapeutic benefits of exposure-based treatment by
pointing out that they already encounter fear situations and yet remain anxious. This
potential objection can be addressed by discussing the differences between naturally
occurring exposure and therapeutic exposure. Table 7.2 lists some of the differences
between natural and therapeutic exposure noted by Antony and Swinson (2000a).
It is important that between-session exposure exercises (i.e., homework assignments) be highly structured and well planned. Antony and Swinson (2000a) note that
individuals with panic disorder may be inclined to carry out exposure on less anxious
240
ASSESSMENT AND INTERVENTION STRATEGIES
TABLE 7.2. Differences between Naturally Occurring Exposure and Therapeutic Exposure
Naturally occurring exposure
Therapeutic exposure
Unpredicted and unsystematic
Predicted, planned, and systematic
Brief duration → perceived defeat
Prolonged duration → perceived victory
Infrequent and sporadic
Frequent and repeated
Threat information exaggerated and safety
information ignored
Threat information evaluated and safety
information is processed
Intolerance of anxiety and heightened anxiety
control efforts
Increased tolerance of anxiety and reduced control
efforts
Reliance on escape and avoidance
Elimination of escape and avoidance
Note. Based on Antony and Swinson (2000a).
days than on days when anxiety is especially elevated. If the exercises are planned in
advance, this will reduce the chance that clients will save homework for their “good
days.”
Within Session versus Between Sessions
Exposure exercises can be conducted with therapist assistance as part of the session
agenda or, more often than not, they are assigned as between-session homework. It
is recommended that the first few exposure exercises be completed with the therapist
present as part of the therapy session. This gives the cognitive therapist opportunity to
observe the client’s response to exposure and correct any problems that might arise.
Select a low to moderately difficult situation so a client’s initial experiences with exposure are successful. The therapist first demonstrates how to carry out the exposure task
(i.e., modeling) and then coaches clients in the correct performance of the task, providing lots of praise and encouragement for confronting their fear and avoidance. In addition the cognitive therapist probes for any automatic anxious thoughts during the exposure demonstration and uses cognitive restructuring strategies to generate alternative
interpretations. In this way a within-session exposure exercise can become an empirical
hypothesis-testing experiment of exaggerated threat appraisals and beliefs.
There are practical reasons for beginning exposure-based treatment with some
therapist-assisted within-session exposure. If the therapist moves too quickly into selfdirected exposure homework assignments, the client might become overwhelmed with
anxiety, resort to escape and avoidance responses, and then give up on the procedure.
There are many pressures on therapists to proceed quickly because often clients have
limited health insurance coverage. Nevertheless, this does not change the risks of introducing self-directed exposure too quickly. Although clients will differ on the amount
of therapist-assisted within-session exposure required in the early phase of treatment,
it would be the rare individual who could proceed directly into self- directed exposure
without requiring at least some practice with the therapist.
Behavioral Interventions
241
Graduated versus Intense Exposure
Most clinicians conduct exposure in a graduated fashion guided by an exposure hierarchy. The hierarchy lists 10–20 situations relevant to the individual’s anxious concerns
that are associated with fear and avoidance ranging from mild to severe intensity. An
expected anxiety level rated on a 0–100 scale is estimated for each situation in the hierarchy. Therapists begin exposure with one of the moderately distressing situations and
proceed as quickly as possible to increasingly more difficult situations (Antony & Swinson, 2000a; Kozak & Foa, 1997). Table 7.3 presents an illustrative exposure hierarchy
that could have been used with Maria in treating her social anxiety.
In this case example the cognitive therapist would begin with a moderately distressing situation such as “walking downtown alone on a busy street” or “meet with
friend at a café” and repeatedly assigns these exposure tasks until there was a significant
reduction in anxiety. Treatment would then progress to the next most distressing situation (e.g., “go shopping with a friend”). Appendix 7.1 presents an Exposure Hierarchy form for use in developing graduated exposure programs for anxious individuals.
Clients rank their experiences from least to most difficult in terms of associated anxiety and avoidance. In addition, individuals are asked to note the core anxious thought
associated with each situation, although this might not be accessible until the individual
initially confronts the situation. Appendix 7.2 is then used to record both within- and
between-session exposure practice sessions. The information from the Exposure Practice Record can be summarized on the Empirical Hypothesis-Testing Form (see Appendix 6.5) and used as a behavioral experiment for evaluating exaggerated threat-related
appraisals and beliefs and their alternative perspective.
TABLE 7.3. Maria’s Illustrative Exposure Hierarchy of Social Situations
Items in fear hierarchy
Sitting at home talking to family
Level of anxiety
(0= no anxiety to 100 =
maximum anxiety/panic)
10
Going for a drive
15
Going for a walk around unfamiliar neighborhood
(minimal risk of meeting a familiar person)
25
Going for a walk around my neighborhood
(greater risk of meeting a familiar person)
35
Walk downtown by myself on busy street
40
Go to movies with a friend
55
Meet with friend in a café
55
Go shopping with a friend
60
Go shopping alone
75
Go grocery shopping alone
80
Go to a party with familiar people
90
Participate in a class or group
Make a speech
95
100
242
ASSESSMENT AND INTERVENTION STRATEGIES
There have been reports of success in using very intensive, massed exposure in
which individuals begin with the most difficult items in the hierarchy. In fact this
ungraded, intensive exposure has been found to be highly successful in treating panic
disorder with agoraphobic avoidance (see discussion by Craske & Barlow, 2001; White
& Barlow, 2002). However, graduated exposure is usually more acceptable to individuals with anxiety disorders who already are concerned about elevated anxiety as a result
of exposure. The prospect of confronting their “worst fears” from the outset is too
risky for most individuals who then might be inclined to refuse further exposure-based
treatment (Antony & Swinson, 2000a). No doubt graduated exposure is the preferred
modus operandi, although the therapist must guard against progressing too slowly up
the exposure hierarchy.
Frequency and Duration
Behavioral manuals on situational exposure recommend daily sessions on a 5-day per
week basis over 3–4 week time intervals with each exposure lasting up to 90 minutes (e.g., Kozak & Foa, 1997; Steketee, 1993, 1999). At its most intense, exposure
procedures have been prescribed 3–4 hours a day, 5 days a week (Craske & Barlow,
2001). Although this latter procedure represents an extreme upper limit, it is probably true that the exposure-based treatments offered in specialized behavioral centers
probably involve more exposure work than what is often seen in more generic naturalistic clinical settings. Failure to achieve within-session and between-session decrements
in fear response with exposure therapy is a significant predictor of poor treatment
response (e.g., Foa, 1979; Foa, Steketee, Grayson, & Doppelt, 1983; Rachman, 1983).
Although a number of factors may be responsible for poor treatment outcome, it is
possible that individuals may have received an insufficient number of exposure sessions
especially when considering the treatment regimens often provided in mental health
centers.
There is some evidence that a concentrated presentation of exposure is more effective than spacing exposure sessions so they occur more sporadically (Antony & Swinson, 2000a; Foa & Kozak, 1985), although there is considerable inconsistency in the
research on this question (see Craske & Barlow, 2001). Antony and Swinson (2000a)
recommend three to six longer practice sessions per week interspersed with brief practices throughout the day. No doubt the most prudent clinical advice would be to encourage at least daily exposure practice when this is a primary intervention strategy in the
treatment plan. Every effort should be made to avoid the negative effects of insufficient
exposure practice on treatment response.
It would appear that prolonged exposure sessions are better than short presentations (Foa & Kozak, 1985), with decreases in anxiety evident after 30 to 60 minutes of
exposure. Foa and Kozak (1986) argue that longer exposure intervals may be necessary
for more pervasive, intense, and complex fears such as agoraphobia. Individual differences in response to exposure can be expected, so the clinician relies on reductions in
subjective anxiety to indicate when to end an exposure session. Antony and Swinson
(2000a) suggest a decrease in anxiety to a mild or moderate level (30 to 50/100) as
indicated by self-report and observer ratings as the criteria for successful completion
of an exposure session. Taylor (2006) considers a 50% reduction in anxiety indicative
of successful exposure. Although differing in their specific findings, the behavioral lit-
Behavioral Interventions
243
erature is clear that frequent, intense, and prolonged exposure is needed to bring about
significant and enduring fear reduction.
Attention versus Distraction
Foa and Kozak (1986) argued that use of distraction strategies that involve cognitive
avoidance such as pretending to be somewhere else, distorting a fear image, concentrating on nonfearful elements of a situation, and generating fear-irrelevant thoughts
or images will diminish encoding of fear-relevant information, impede fear activation,
and so lead to failure in emotional processing. Thus it is recommended that clients fully
attend to the fear elements of a situation during exposure and to minimize distraction
as much as possible (Craske & Barlow, 2001).
The empirical research on the effects of attention versus distraction in exposurebased treatment has not been consistent (for reviews, see Antony & Swinson, 2000a;
Craske & Barlow, 2001). The best conclusion is that distraction may not have a particularly negative effect in the short term but it does appear to undermine treatment effectiveness in the long term. Based on Antony and Swinson (2000a), we make the following
recommendations for enhancing the effectiveness of exposure:
1. Instruct clients to fully attend to the fear elements of the situation or image.
This is accomplished by having clients verbally describe elements of the situation, their
reaction to these features, and their interpretations of what they see or feel. Taylor
(2006) notes that the intensity of the exposure experience can be adjusted by altering
the amount of detail the client describes in the fear situation.
2. Minimize overt and covert sources of distraction as much as possible. Frequently
ask clients what they are thinking about at this moment. Remind clients to refocus on
the task at hand if attention becomes distracted.
3. Encourage clients not to fight their anxiety by trying to suppress their feelings.
Antony and Swinson (2000a) note that efforts to suppress anxious feelings or even
the attempt to reduce discomfort could paradoxically maintain or increase discomfort.
Thus “accepting the fear” is probably the most beneficial attitude to maintain during
exposure.
Controlled Escape versus Endurance
Standard exposure-based protocols assume that clients should continue (i.e., endure)
with an exposure exercise until there is a significant reduction in anxiety (e.g., Foa &
Kozak, 1985). An alternative view is that exposure should continue until individuals feel
their anxiety level is “too high” or intolerable, at which point they can escape from the
situation as long as there is an immediate return to the fear situation a few minutes later
(Craske & Barlow, 2001).
If one adheres to a behavioral view of anxiety reduction, then endurance is the preferred method in order to ensure within-session habituation of anxiety (Foa & Kozak,
1986). On the other hand, if anxiety reduction is explained in terms of increased selfefficacy or the incorporation of safety signals, then controlled escape would be permissible (Craske & Barlow, 2001). Once again the empirical research is not entirely consistent on this issue (see review by Craske & Barlow, 2001). From a cognitive perspective,
244
ASSESSMENT AND INTERVENTION STRATEGIES
controlled escape may be problematic because it could reinforce beliefs that the situation is fraught with danger, high anxiety is intolerable, and the best response is escape.
For these reasons we believe that encouraging clients to endure exposure sessions until
there is a significant reduction in anxiety will provide the best disconfirmatory evidence
against exaggerated appraisals of threat and personal vulnerability.
Collaboration and Client-Oriented Control
Perceived predictability and control are important for individuals engaged in exposurebased treatment (Antony & Swinson, 2000a). Consistent with the cognitive therapy
orientation, there should be a strong collaborative atmosphere, with clients directly
involved in setting their exposure homework assignment. Individuals should be assured
that they will never be asked to do something they don’t “want” to do and that the pace
of the exposure treatment is under their own control. Naturally the therapist will be
encouraging clients to challenge themselves, but there should be no hint of a coercive or
heavy-handed approach. Some cognitive restructuring may be necessary before a reluctant client agrees to undertake some aspect of the exposure hierarchy. It may also be
useful to ask the client for an expected timetable for progressing through the hierarchy.
That way the therapist can correct any faulty expectations about speed of progress in
light of the client’s actual pace of exposure treatment.
Antony and Swinson (2000a) noted that some exposure situations will be inherently
unpredictable such as social situations (e.g., the socially anxious client asked to initiate a
brief conversation with work colleagues). In such cases the therapist might have to work
on preparing the client for possible negative outcomes. At other times one might want to
build some unpredictability into later exposure exercises so the client is better prepared
to handle all the vicissitudes inherent in naturalistic daily life experiences.
Safety Signals and Partner-Assisted Exposure
Most behavioral therapists recommend that reliance on safety signals be eliminated
during exposure (e.g., Taylor, 2000: White & Barlow, 2002). Some of these behaviors
can be quite subtle such as the production of automatic responses like tensing or holding
one’s breath. Dealing with safety cues during exposure means that the therapist must
first identify these responses, wean clients off the safety signals by building this into the
exposure exercises, and encouraging the client to refrain from safety seeking (Taylor,
2000). Eliminating safety signals is important in therapy, because their continued presence is a form of avoidance that undermines disconfirmation of the threat and vulnerability beliefs. In the illustrative case example, Maria believed that maintaining a neat
and tidy appearance would guarantee protection against the negative evaluation of others. This served a safety-seeking function that was targeted in therapy through cognitive
restructuring conducted concurrently with social situation exposure assignments.
In some anxiety disorders, like agoraphobia, a particular family member or friend
may be a powerful safety cue for the anxious client. When reviewing exposure homework, the therapist must always inquire whether the task was completed alone or with
partner assistance. If there is excessive reliance on a partner, this should be built into the
exposure hierarchy so that clients are gradually weaned off their dependence on others
as they progress up the hierarchy. Individuals who can not venture into an anxious situ-
Behavioral Interventions
245
ation without support of a friend, family member, or spouse are unlikely to maintain
long-term gains in anxiety reduction (Antony & Swinson, 2000a).
Anxiety Management during Exposure
Given the importance of frequent and prolonged exposure to fear stimuli, one might
assume that any form of anxiety management has no place in exposure-based treatment. Is it not better that the client remains in a heightened state of anxiety so that
the full effects of the disconfirming evidence can be processed and a natural reduction
in anxiety is achieved? In most instances it would be better to refrain from deliberate
anxiety management. However, there are times when some anxiety management may
be necessary in order to encourage prolonged and repeated exposure to high anxietyprovoking situations. For example, clients who experience extreme levels of anxiety in
a wide range of situations or others who have exceptionally low tolerance for anxiety
could be taught some anxiety management strategies to reduce anxiety to the moderate
range, which is more optimal for successful exposure.
Steketee (1993) describes four types of anxiety management strategies that can be
used in exposure-based treatment to reduce subjective anxiety. The first is cognitive
restructuring in which individuals challenge their exaggerated threat appraisals by noting evidence in the exposure situation that the danger is not as great as they expect
and that anxiety eventually declines naturally. Beck et al (1985, 2005) list a number of
“coping statements” that can be used by clients to encourage endurance in the anxious
situation. The aim of these cognitive strategies is to alter the appraisals and beliefs
responsible for the elevated anxiety in the situation. With Maria, cognitive interventions
focused on her erroneous beliefs about the source of her anxiety (e.g., “that other people
are looking at me”).
A second anxiety management approach is to provide the client relaxation training
such as progressive muscle relaxation, controlled breathing, or meditation. These coping responses could then be used during exposure to reduce anxiety. However, Steketee
(1993) warns that relaxation has been shown not to be particularly effective in moderate to high anxiety. Also relaxation could easily be transformed into an avoidance or
safety-seeking response. For these reasons, relaxation training is rarely incorporated
into exposure-based treatment. Occasionally, however, it could be taught as a means
of bolstering perceived control for anxious individuals who initially refuse exposure
intervention because of low self- efficacy expectations. In other cases, like with Maria,
reliance on controlled breathing can prove detrimental because her breathing rate was
so exaggerated during peak anxiety that it actually bordered on hyperventilation and
probably drew attention from others.
A third approach is to use paradoxical intention in which a person is instructed to
exaggerate her anxious response in a fear situation. Asking people to exaggerate their
fear often highlights the absurdity and improbability of the fear, which has the intended
paradoxical effect of causing a reevaluation of the actual threat and vulnerability associated with the situation (Steketee, 1993). For example, a person with panic disorder
and agoraphobic avoidance might be reluctant to take a walk five blocks from home.
Assuming proper medical clearance was obtained, the person could be instructed to jog
when he feels intensely panicky from an accelerated heart rate. The jogging, of course,
would elevate the heart rate even further but it would cause its reattribution to increased
246
ASSESSMENT AND INTERVENTION STRATEGIES
physical activity. This would probably result in a reduction of subjective anxiety to a
more tolerable level.
A final anxiety management strategy involves calling the therapist, a family member, or a friend for reassurance and support (Steketee, 1993). Given our previous discussion on safety seeking, this form of intervention could quickly undermine the effectiveness of exposure and so should be used sparingly. Any evidence that this form of support
seeking has become an entrenched coping style would require that it be immediately
faded from treatment. On the other hand, it may be that the provision of some support
may be needed for a brief interval, especially in the early phase of treatment, to encourage participation in the exposure sessions. Beck et al. (1985, 2005) recommended the
use of significant others to serve as auxiliary therapists in carrying out behavioral exercises. White and Barlow (2002) concluded from their review of the empirical literature
that attending to the client’s social support system and utilizing significant others in
homework assignments might actually enhance the effectiveness of exposure treatment,
especially for individuals with agoraphobia. In the early stage of treatment, family members accompanied Maria to long avoided social situations but their presence was quickly
faded as soon as possible. At the very least, then, the role of partners, family, and close
friends should be considered when setting between-session exposure assignments.
Clinician Guideline 7.2
Effective exposure interventions must activate fear schemas and provide disconfirming
threat information that will result in modification of the client’s fear structure. This is best
accomplished by providing frequent, moderately intense, and prolonged within-session and
between-session exposure that is implemented in a planned, systematic, and graduated manner. Clients should be given a cognitive rationale for the exercises with a therapeutic orientation that emphasizes exposure as a direct, experiential evaluation of anxious appraisals and
beliefs. To enhance exposure assignments safety seeking, distraction, and escape/avoidance
should be eliminated. Clients should engage in daily exposure between sessions.
Situational (In Vivo) Exposure
The most common form of exposure-based treatment involves repeated, systematic presentation of real-life experiences (Craske & Barlow, 2001). We see situational or in vivo
exposure used most often with specific phobias, panic disorder with agoraphobic avoidance, OCD, and social phobia. In such cases the exposure hierarchy consists of a range
of real-life situations that elicit varying degrees of avoidance. Taylor (2006) notes that
exposure should not be used if the client has poor impulse control, uncontrolled substance use disorder, suicidal ideation or urges, or engages in stress-induced self-injurious
behavior. Furthermore, clients should have a physical examination by a physician to
determine if there are any medical contraindications for engaging in certain types of
exposure interventions.
As discussed previously, exposure is introduced as a powerful “learning through
experience” intervention that can reduce anxiety. However, the therapist will have to
take special consideration of clients who had a past negative experience with exposure.
Behavioral Interventions
247
Antony and Swinson (2000a) suggest that the therapist focus on highlighting the differences between “bad” exposure and “good” exposure (see Table 7.2). In the end the
therapist must provide a convincing rationale for exposure that will encourage the client’s full participation in the exposure procedures.
When implementing exposure, begin with therapist-assisted demonstrations in the
treatment session followed by well-planned, structured, and graduated between-session
self-directed exposure assignments that evoke moderate anxiety. Exposure should be
done daily with many of the sessions at least 30–60 minutes long and continued until
there is a 50% reduction in subjective anxiety. Each session begins with a 0–100 rating of initial anxiety level and recording any anticipatory anxious thoughts about the
exposure task. The individual then enters the fear situation and provides an anxiety
rating every 10–15 minutes. In addition clients should take note of any specific anxiety
symptoms experienced during the exposure session and their interpretation of the symptoms. As well, any apprehensive thoughts or images should be noted and clients should
be encouraged to use cognitive restructuring strategies to correct their thinking. A final
anxiety rating is completed at the end of the exposure session and observations noted
about the outcome of the exposure session. One of the core beliefs targeted in Maria’s
exposure assignments was “People are looking at me and will notice that I am anxious,
that I can’t breath, and conclude there is something wrong with me.”
The postexposure evaluation session is perhaps the most important part of the
intervention from a cognitive perspective (see previous chapter on consolidation and
summary stages of behavioral experiments). The cognitive therapist reviews in detail the
Exposure Practice Form and other materials that document the client’s thoughts, feelings and behavior during the exposure exercise. In cognitive therapy, exposure is viewed
as a behavioral experiment or empirical hypothesis-testing exercise. Thus the client’s
observations of the exposure exercise can be recorded on the Empirical HypothesisTesting Form (see Appendix 6.5) and this can be used to emphasize those features of
the exposure experience that disconfirmed core anxious appraisals and beliefs. It is
expected that repeated evaluation of multiple exposure experiences will ultimately provide the disconfirming evidence needed to modify the client’s anxious thoughts and
beliefs and lead to long-term reduction in anxiety. Examples of graded in vivo exposure
can be found in various behavioral treatment manuals (e.g., Antony & McCabe, 2004;
Kozak & Foa, 1997; Foa & Rothbaum, 1998; Steketee, 1993), as well as in Chapter 6
on empirical hypothesis testing.
Clinician Guideline 7.3
In vivo exposure is perhaps the most powerful behavioral intervention for fear reduction.
Whenever possible, employ this therapeutic tool in the treatment of anxiety disorders.
Imaginal Exposure
The goal of any exposure intervention is to provoke anxiety or distress and allow it to
decrease spontaneously without recourse to avoidance, neutralization, or other forms
of safety seeking. There is considerable empirical evidence that this objective can be
248
ASSESSMENT AND INTERVENTION STRATEGIES
achieved with imaginal exposure, although most behavior therapists recommend the
use of in vivo exposure whenever possible because it appears to yield more potent and
generalizable treatment effects (e.g., Antony & Swinson, 2000a; Foa & Kozak, 1985;
Steketee, 1993). Foa and McNally (1996) stated that imaginal scripts can not be as
effective as real-life exposure, because they provide improvished informational input
and so are less evocative of the fear memory structure. However, there are times when
imaginal exposure is the preferred modality because in vivo exposure is impractical (or
impossible), or the addition of imaginal exercises enhances treatment maintenance of
externally based exposure (Kozak & Foa, 1997). The following is a list of occasions
when imaginal exposure might be the more appropriate therapeutic modality.
•฀ When the object of fear is a thought, image, or idea, imaginal exposure may be
the only possible therapeutic approach (e.g., in OCD thinking of the end of the
world, of eternal damnation, of committing the “unpardonable sin”).
•฀ Imaginal exposure is used when it is impractical or unethical to utilize in vivo
exposure (e.g., fear of shouting obscenities in church, thoughts of accidentally
causing harm or injury to another, fear of natural disasters).
•฀ In PTSD imaginal exposure is often utilized when fear is associated with memory
of a trauma that happened in a distant geographic location or at an earlier time
of life (Keane & Barlow, 2002).
•฀ Borkovec (1994) has argued that worry is a conceptually based cognitive strategy used to avoid aversive imagery and the physiological arousal associated with
threatening topics. Imaginal exposure has become an important component of
CBT protocols for GAD (Brown, O’Leary, & Barlow, 2001; Rygh & Sanderson,
2004).
•฀ Imaginal exposure is effective as a preparatory skills exercise such as in treating
public speaking anxiety where imagery and role-play rehearsal are utilized for
skills acquisition prior to in vivo exposure.
•฀ Finally, imaginal exposure may be employed initially when a client refuses to
engage in real-life exposure in order to facilitate the eventual acceptance of in
vivo exposure exercises (Antony & Swinson, 2000a).
Implementation
The general guidelines previously discussed under situational exposure are applicable to
imaginal exposure, although the following caveat should be taken into account. First,
flooding or abrupt exposure procedures, which involve the immediate presentation of
the most feared scenario, are used more often in imaginal than in in vivo exposure. This
is particularly true for the imagery exposure used in PTSD or GAD where a hierarchical approach to trauma or “worst-case scenario” may not be necessary. Since flooding
is more efficient and equally (or more) effective to hierarchical exposure (Foa & Kozak,
1985; White & Barlow, 2002), clinicians should consider whether an intensive form of
imaginal exposure can be applied.
Second, imaginal exposure sessions are usually no more than 30 minutes and so
are much shorter in duration than situational exposure. Sustained imagery exercises
require a great deal of attentional resources so most individuals would not be able to
maintain their full concentration on the imagery task for prolonged periods. However,
Behavioral Interventions
249
it is likely that the number of imaginal exposure sessions is no more or less than for in
vivo exposure.
Third, cognitive avoidance is more difficult to control in imaginal than real-life
exposure sessions (Foa & Kozak, 1986). Individuals can distract themselves from the
fear image by replacing it with another thought or image, or they can imagine less
threatening versions of the fear scenario. This will weaken the effectiveness of exposure
by undermining the image’s capacity to activate fear schemas (see Foa & McNally,
1996).
To overcome this inherent limitation with symbolic representation, behavior therapists have introduced certain modifications in order to enhance the effectiveness of
imaginal exposure. One procedure is to require the client to write down a full description of the fear imagery script (e.g., Kozak & Foa, 1997; Rygh & Sanderson, 2004).
(See discussion on imaginal reprocessing and expressive writing in Chapter 6.) For the
scripted narrative to be effective it must include details that have emotional significance
to the client as well as the client’s anxiety response (e.g., increased tension, heart palpitations) to the fear scenario (Kozak & Foa, 1997). Developing an effective fear narrative
can be difficult, so this is usually done in the session with the therapist using guided
discovery to help the client come up with an effective imagery script. Once a script has
been developed, the first imaginal exposure sessions should be conducted in the therapy
session. The exposure exercise begins by having the client read the narrative aloud and
then closing her eyes to generate a full and complete image of the fear scenario. If the
image starts to fade, the client should open her eyes and reread sections of the narrative
to reestablish the image. This process continues for the duration of the exposure session. After repeated presentations of the fear imagery, it may be necessary to modify the
narrative in order to maintain its evocative properties. The following is an example of
a narrative script for a 55-year-old man with GAD who was terrified of financial ruin
even though he had attained a high level of financial security.
“You wake up on a Thursday morning feeling particularly anxious. You’ve had very
little sleep because you’ve been tossing and turning all night long, worried about
your finances. You finally crawl out of bed feeling tired, exhausted. You have a lowgrade headache, your muscles ache, and you can hardly walk as you shuffle to the
kitchen. The house feels quite cool as you are the first up on this particular morning. It is dark and dreary outside with a light rain spattering on the window pane.
You sit at the kitchen table, your mind continuing to race about your investments
and whether you made the right decision while doing some online trading. You have
a sickening feeling that you left yourself financially vulnerable by overinvesting in
that tech stock. You notice that you are feeling tense, your chest aches, and your
heart is racing. You try to get control but the more you try the worse it gets. You are
now convinced that you’ve made a terrible mistake. How could you be so stupid as
to invest so much money in a high-risk stock? You can feel yourself becoming more
and more agitated, you get up and start pacing, wringing your hands as you walk.
All you can think about is that stupid investment when suddenly you notice that the
mail has come for that day. You try to distract yourself by going to the mailbox.
There is quite a bite of correspondence but your eye immediately drops to an envelope from your bank. You notice that it is from your discount brokerage firm. You
know this is the monthly statement of your investments. With trembling hands, and
250
ASSESSMENT AND INTERVENTION STRATEGIES
a feeling of nausea in your stomach you tear open the envelop. Your eyes immediately fix on the monthly balance. You can’t believe what you see; your investments
have been practically wiped out! A couple of important investments have gone sour
and your hard-earned investments have been decimated. You feel your legs weaken,
your hands are shaking, and you think you are going to be sick. You drop into the
chair, your heart feels like it is going to explode, and you feel sharp chest pains. You
can’t believe what you see and so you keep looking at the numbers. And yet, there
it is; you’ve lost thousands and thousands of dollars. You realize you are finished,
your investment portfolio is ruined. What will you do now?”
Another procedure that has been introduced to enhance imaginal exposure is audio
habituation training. A recording of the fear scenario is made on a CD so that the fear
script is presented repeatedly without interruption. The client is instructed to listen to
the CD and to get into the scenario depicted as fully as possible. The CD is allowed to
play repeatedly for 20–30 minute exposure sessions. It is important that clients make
the CD recording themselves so that they are listening to their own voice. A number of
single-case reports have described the effectiveness of audiotaped exposure for obsessional fears in which the audiotape not only enhances the imaginal exposure experience
but reduces the opportunity for clients to engage in covert neutralizing responses that
would undermine the exposure experience (e.g., Headland & McDonald, 1987; Salkovskis, 1983; Thyer, 1985).
Clinician Guideline 7.4
Imaginal exposure is particularly useful in the treatment of OCD, GAD, and PTSD where
the source of anxiety is a thought, image, or memory. Abrupt forms of exposure or flooding
are more often used along with narrative scripts or audiorecordings of the imaginal fear to
ensure sufficient fear activation and reduction in cognitive avoidance.
Exposure to Bodily Sensations
Certain physical sensations such as chest pain, shortness of breath, dizziness, nausea,
and the like can elicit, or at least further exacerbate, anxiety because they are erroneously misinterpreted in a threatening manner. This catastrophic misinterpretation of
bodily sensations is especially characteristic of panic disorder (Beck, 1988; Beck &
Greenberg, 1988; D. M. Clark, 1986a). As with any fear stimulus, it is important that
clients experience repeated exposure to their anxiety-provoking bodily sensations. This
is accomplished by conducting various “panic induction exercises” that involve deliberate activation of bodily sensations such as overbreathing or hyperventilating, breathing
through a straw, running on the spot, and so on. In cognitive therapy the purpose of
these exposure exercises is to activate fear schemas, in this case fear of bodily sensations, and provide anxious individuals with experiences that correct their erroneous
symptom equation (e.g., that chest pain = elevated risk of heart attack; Beck & Greenberg, 1987).
Exposure to bodily sensations in cognitive therapy bears some resemblance to
Barlow’s interoceptive exposure that involves repeated reproduction and exposure to
Behavioral Interventions
251
uncomfortable arousal-related bodily sensations (White & Barlow, 2002; Taylor, 2000).
The purpose of interoceptive exposure is fear reduction of specific bodily cues through
repeated exposure (Craske & Barlow, 2001). However, in cognitive therapy these exercises are used differently to activate the fear schemas associated with bodily sensations
and provide corrective evidence against the catastrophic misinterpretations of physical
symptoms. Although interoceptive exposure is most often used in cognitive therapy for
panic disorder, it is relevant for any anxious individual who fears a particular body sensation (Antony & Swinson, 2000a). A more detailed account of this type of exposure
can be found in the next chapter on panic disorder.
Clinician Guideline 7.5
Use exposure to bodily sensations to activate the client’s fear schema by intentionally producing the body sensations associated with anxiety in order to provide corrective evidence
against the catastrophic misinterpretation of the sensation. The procedure is used most frequently in the treatment of panic disorder.
RESPONSE PREVENTION
Response prevention involves the deliberate suppression of any coping strategy, such as
a compulsion, neutralization, or other control response performed to alleviate anxiety
or discomfort (D.A. Clark, 2004). As a behavioral intervention, response prevention
is most often used in conjunction with exposure interventions, especially in the treatment of OCD. However, when viewed more broadly as the prevention of maladaptive
coping responses that contribute to the persistence of anxiety, response prevention can
be an important treatment component for any of the anxiety disorders. For instance,
with Maria it was important to reduce her reliance on “controlled” breathing when she
became anxious because it actually intensified her anxious state.
Response prevention is most relevant for addressing the deliberate safety-seeking
strategies that anxious individuals employ during the elaborative phase of anxiety (see
Chapter 2, Figure 2.1). In Chapter 5 we listed 34 behavioral and emotional coping
responses that might be used to neutralize anxiety (see Appendix 5.7). Moreover, highly
anxious individuals often engage in effortful cognitive strategies aimed at alleviating
discomfort such as deliberate thought suppression, rationalization, and the like (see
Appendix 5.9). Response prevention, then, is a robust intervention strategy designed to
eliminate problematic behavioral, that is, emotional and cognitive responses that lead to
premature termination of exposure to a fear stimulus.
In essence any therapeutic intervention that seeks to suppress the expression of
safety-seeking responses in the context of anxiety arousal is a form of response prevention. The goal is to help clients become more aware of their maladaptive coping
responses, suppress these responses, and engage in more adaptive responses to ensure
continued exposure to the fear-eliciting situation. Initially the therapist can model
response prevention in the therapy session and then proceed to coaching the client in
similar coping strategies. The eventual goal is for the client to engage in self-directed
response prevention in the naturally occurring anxious situation.
252
ASSESSMENT AND INTERVENTION STRATEGIES
Implementing Response Prevention
There are a number of steps involved in implementing response prevention. (See also
Rygh & Sanderson, 2004, for a description of response prevention for GAD.)
Identify Maladaptive Coping and Neutralization
In order to implement response prevention, the therapist must first identify the cognitive, behavioral, and emotional responses used to terminate exposure to fear stimuli and
reduce anxiety. The assessment forms in Appendixes 5.7 and 5.9 are quite useful for
this purpose. In addition direct observation of the client during exposure to fear situations may identify other more subtle automatic safety-seeking responses that should be
addressed in the exposure and response prevention sessions. For example, Maria would
frequently interrupt therapy sessions by asking the therapist if he thought she looked
alright and she would also frequently leave her seat in order to check on her appearance
in the mirror. Response prevention of these safety-seeking responses and their underlying beliefs was an important part of the therapy. In most cases it is helpful to assign
homework that requires self-monitoring of safety-seeking and other coping responses
in order to heighten the client’s awareness of these strategies. Repeated self-monitoring
of one’s anxiety responses and control efforts can help bring fairly automatic processes
under more elaborative, conscious control.
Provide Rationale for Response Prevention
It must be explained to clients why the prevention of maladaptive coping responses
is a critical component of cognitive therapy. Often the rationale for response prevention is presented when educating the client about exposure-based interventions. Rygh
and Sanderson (2004) suggest that a cost–benefit approach can be used in which the
short-term anxiety reduction associated with maladaptive coping and neutralization
responses is offset by the long-term persistence of anxiety. It should be explained that
long-term anxiety reduction will only occur when the underlying faulty appraisals and
beliefs about exaggerated threat and vulnerability are truly modified. The most effective
strategy for changing these attitudes is learning to tolerate anxiety and then letting it
decline naturally.
Preventing maladaptive responses that prematurely terminate anxiety is an important part of this treatment approach. The following is a clinical excerpt that illustrates
educating a client with panic disorder on the importance of response prevention.
THERAPIST: Derek, I notice from the behavioral checklist [Appendix 5.7] that whenever
you feel anxious about chest pains you immediately stop all activity, rest, and try to
control your breathing in an effort to relax yourself.
CLIENT: Yeah, I’ve done this for so long it is kinda automatic now. I keep thinking it is
really important to relax and get control of myself.
THERAPIST: I also notice that on other occasions, when the anxiety gets really bad, you’ll
look on the Internet for medical information, make an appointment with your family doctor, or even go to the hospital emergency department. These all look like
ways of seeking reassurance that you are alright and not having a heart attack.
Behavioral Interventions
253
CLIENT: I’ve been doing these things for years but the anxiety seems to always come
back.
THERAPIST: Derek, that’s an important observation that you just made. So you find that
trying to relax or seeking medical advice calms your anxiety for a while but then it
comes back just as strong as ever.
CLIENT: That’s exactly what happens.
THERAPIST: It’s kind of like the old adage in reverse “short-term gain but long-term
pain.” Responses like trying to relax or seeking reassurance may work in the short
term but over time they actually contribute to the persistence of anxiety. They prevent you from learning to tolerate anxiety and that nothing terrible will happen to
you because of the physical symptoms of anxiety. By artificially cutting short the
anxiety, it doesn’t have a chance to disappear naturally and you never have a chance
to learn that your fearful thoughts about chest pain and heart attacks are based on
exaggerated misunderstandings about risks to your health.
CLIENT: Are you saying that going to the doctor or trying to relax are bad, that these
things actually make me more anxious?
THERAPIST: Yes, that is exactly what I am saying. These coping strategies prevent you
from actually learning ways to deal with faulty beliefs about risks to your health.
And so the anxiety you have over chest pain and heart attacks continues unabated.
You recall that earlier we talked about the exposure exercises as an important way
to learn how to let anxiety decline naturally. It is also very important to eliminate
some of these coping strategies like rest, trying to relax, or seeking medical reassurance that artificially interrupt the anxiety response. So while you are doing the
exposure exercises, I would also like to work with you on ways to reduce these
problematic coping behaviors. We use procedures called response prevention which
focus on suppressing certain maladaptive responses. Would you like to look at some
strategies we could use to reduce or even eliminate these problem behaviors and
build up better responses to the anxiety?
CLIENT: Sure, this sounds like a good idea.
Prepare Client for Heightened Anxiety
Individuals must be forewarned to expect an immediate increase in anxiety with prevention of safety-seeking responses and prolonged exposure to fear stimuli. Although
individuals will differ in the duration of peak anxiety, some reduction in anxiety should
be evident after 30–60 minutes of exposure. However, continued prevention of maladaptive coping and neutralization may be necessary for several hours after an exposure
session. For example, individuals with obsessive fears of contamination can engage in
washing and cleaning rituals that take hours to complete. In such cases the response prevention phase of an exposure homework assignment can extend over a 2–3 hour time
period. The approximate duration of a response prevention session should be discussed
with clients. Most often clients will be encouraged to continue with their response prevention until their anxiety reaches the mild range.
There are times when an individual’s anxiety is so intense over a prolonged period
of time that she refuses to engage in exposure and response prevention. In such cases
254
ASSESSMENT AND INTERVENTION STRATEGIES
certain anxiety management strategies such as distraction, controlled breathing, and
progressive muscle relaxation can be temporarily introduced. It is important that these
strategies be employed briefly because they can interfere with full exposure to the fear
stimuli. In the end the effectiveness of exposure and response prevention will be weakened if clients continue to rely on anxiety management. This would suggest that the
individual’s exaggerated threat appraisal of the physical symptoms of anxiety remains
intact.
Instruct Client about “Blocking” Strategies
A number of strategies can be used to suppress maladaptive coping behavior and
other forms of safety-seeking responses. First, the client can write down a list of selfinstructional coping statements that can be used as reminders of the benefits of preventing maladaptive responses and the costs of continued reliance on problematic safetyseeking behavior. Second, individuals could develop a repertoire of competing activities
that interfere with performance of the maladaptive coping behaviors. For example, individuals who hold their breath when anxious could practice diaphragmatic breathing
or those who tend to overbreathe when anxious could focus on holding their breath
between exhalations. To compete with compulsive checking, an individual could immediately leave the situation so that repeating a check becomes more difficult. For instance,
Maria was restricted to using mirrors only at certain times of the day and to refrain from
carrying a mirror in her purse. Considerable exploration will be necessary to develop a
repertoire of competing activities that would effectively block safety-seeking behavior. It
is likely that these competing responses will be quite idiosyncratic to the individual and
the specific safety-seeking behavior under consideration.
A third response prevention strategy that is probably the most effective in blocking
problematic coping responses is paradoxical intention. This involves having the client
engage in behaviors that are completely opposite to the safety-seeking response. For
example, a person who tries to rest whenever he feels anxious for fear that his pulse
rate is too high could engage in a high- energy physical activity when he feels anxious.
Someone who uses cognitive avoidance or distraction to deal with her anxiety could be
instructed to fully attend to the fear stimulus. And of course the person who relies on
escape and avoidance would be encouraged to remain in the fear situation. The client
who suppresses anxious feelings would be instructed to openly express his emotions,
whether they be fear or anger. It is likely that the deliberate performance of a behavior
that is opposite to the coping response will provide the most effective response prevention.
And finally, the support and encouragement of family and friends can be a powerful incentive to refrain from problematic responses to anxiety. With proper instruction from the therapist family members can serve as “coaches” to encourage exposure
and response prevention. Given Maria’s excessive reassurance seeking, family members
would need to be instructed on how to handle her requests for reassurance about her
physical appearance. Of course, involvement of significant others has to be monitored
carefully so that the person does not become a safety cue. Furthermore, the therapist
should provide verbal encouragement and be available by phone between sessions to
assist clients who might have difficulty blocking their maladaptive coping responses.
Behavioral Interventions
255
Develop Alternative Coping Responses
The adaptive alternative that is promoted in cognitive therapy is continued exposure to
the fear stimulus. Any coping responses that encourage the client to wait for anxiety
to dissipate naturally is considered an adaptive approach. For example, a client with a
long-standing panic disorder was very terrified of panic attacks. The core belief was she
might lose control and eventually go insane. Any signs of anxiety, especially trembling,
shaking, or crying, were misinterpreted as loss of control. She responded by tensing her
muscles, distracting herself, and trying to suppress her anxious feelings. To counter these
futile attempts at anxiety control, a form of paradoxical response prevention was formulated. Whenever she noticed the first signs of anxiety, she was to go to her bedroom,
stand before a full-length mirror, and purposefully shake and cry as hard as possible.
She was to watch herself do this in the mirror until her anxiety level dropped significantly. This plan for coping with anxious episodes served several functions. It encouraged direct exposure to the physical symptoms that frightened her. It also blocked her
maladaptive coping responses and it usually ended with a good laugh, which initiated
an emotional state contrary to anxiety. In sum, effective response prevention should not
only specify the safety-seeking responses that should be blocked or suppressed, but also
alternative ways of responding that promote adaptive exposure.
Challenge Problematic Cognitions
The cognitive therapist is always attentive to any faulty thoughts or beliefs that might
lead to continued reliance on safety-seeking responses and undermine response prevention. This can be done by questioning clients on their automatic thoughts about perceived need to avoid or control anxiety as well as by examining self-monitoring records
for maladaptive safety-seeking cognitions that occurred during exposure assignments.
Once such thinking is identified, cognitive restructuring can be employed to modify the
anxious appraisals and beliefs (see Chapter 6).
Certain themes are common in the automatic thoughts and beliefs that maintain
safety seeking and interfere with response prevention. These include an intolerance of
anxiety and uncertainty, a need to maintain control, the importance of minimizing risk,
and the maintenance of safety and security. Individuals with anxiety will often express
beliefs like “I can’t stand the anxiety,” “I need to be certain that I haven’t left the stove
burners on and could cause a fire,” “If I don’t maintain strict control over my emotions,
people will notice there is something wrong with me,” “I can’t stand to take risks; it’s
better to be safe than sorry,” “The more I feel peace and comfort the better my physical
and mental health,” or “If I look perfect, I can avoid the negative evaluation of familiar
people [Maria].” In many cases response prevention of maladaptive coping and safety
seeking will not be accepted as long as the anxious person endorses this way of thinking. Thus the cognitive therapist should probe for problematic cognitions whenever clients fail to follow through on response prevention.
Record and Evaluate
As with any intervention, it is essential that clients maintain some record of their response
prevention efforts between sessions. The Response Prevention Record in Appendix 7.3
256
ASSESSMENT AND INTERVENTION STRATEGIES
can be used for this purpose. This form can be completed when clients engage in exposure homework assignments or when they prevent maladaptive coping during spontaneous, naturally occurring anxiety episodes. Although the form collects data on anxiety
levels and urge to engage in the “prevented response,” the cognitive therapist should
always probe for clients’ cognitions about response prevention and safety-seeking behavior when reviewing the form.
Clinician Guideline 7.6
Preventing maladaptive coping behavior and other forms of safety-seeking responses is an
important component of cognitive therapy that promotes exposure to information that disconfirms the client’s faulty threat and vulnerability beliefs.
DIRECTED BEHAVIORAL CHANGE
As previously discussed, individuals with anxiety disorders often exhibit problematic
behaviors that require modification or they may present with behavioral deficits that
actually contribute to their anxious state. An individual with social phobia may have
performance deficits in interpersonal and communication skills, although Antony and
Swinson (2000b) remind us that most people with social anxiety have better interpersonal skills than they think. However, social behavioral performance deficits may also
be evident in other anxiety disorders. It can occur in the person with panic disorder
and agoraphobic avoidance who has shunned social settings for many years, or the
individual with chronic OCD who might avoid others because of obsessions of doubt
or contamination. Moreover, individuals with PTSD often have significant social withdrawal and other interpersonal difficulties (Turner, Beidel, & Frueh, 2005). In such
cases a skills-training component might be included in the treatment plan.
Directed behavioral change refers to intervention strategies that teach individuals how to change specific behaviors in order to improve their personal effectiveness
at home, at work, and in interpersonal relations. In the anxiety disorders behavioral
change strategies typically focus on improving prosocial skills, assertiveness, or verbal
and nonverbal communication (see Antony & Swinson, 2000a, 2000b, for further discussion). Table 7.4 presents the steps normally involved in behavioral change interventions.
TABLE 7.4. Therapeutic Elements in Direct
Behavioral Change Interventions
•฀ Didactic instruction or psychoeducation
•฀ Modeling specific behaviors
•฀ Behavioral rehearsal
•฀ Corrective feedback and reinforcement
•฀ In vivo homework assignments
•฀ Self-monitoring and evaluation
Behavioral Interventions
257
When initiating a behavioral change intervention, the therapist begins with didactic instruction aimed at preparing the client for behavioral rehearsal. Goldfried and
Davison (1976) comment that this didactic introduction is necessary for ensuring that
the client recognizes that behavioral change is needed, to accept behavioral rehearsal
as an important step in learning new behaviors, and to overcome any anxiety about
role playing. In addition, the therapist provides specific information that helps clients
learn the difference between their maladaptive behaviors and more effective prosocial
behaviors.
In cognitive therapy a rationale should be given for shifting therapy from a focus
on the cognitive basis of anxiety to this more behavioral orientation. Clients should be
informed that these interventions are not intended as a direct anxiety-reduction strategy, but rather their aim is to improve one’s functioning and confidence in social situations. Improved social functioning might have an indirect anxiolytic effect by increasing
the frequency of positive responses from others, which in turn would increase a person’s
motivation to expose himself to anxiety-provoking encounters with others.
Modeling plays an important role in teaching anxious clients how to engage in
more effective interpersonal behavior. The therapist demonstrates the skill that is to
be learned and then discusses with the client how to perform the behavior in question.
Even though didactic explanations of new behaviors are important, nothing can substitute for actually showing a client how to respond. For example, a person with social
anxiety had a tendency to talk too quickly when conversing at work. Even though it
ensured quicker escape from an anxious social interaction, it interfered in the quality of
her communication and actually intensified her subjective anxiety. This acceleration of
her speech actually occurred in the therapy session. The therapist was able to interrupt
the conversation, point out that her speech was accelerating, and then demonstrate a
more appropriate rate of speech. This modeling led naturally into the next phase of the
behavioral change intervention.
Behavioral rehearsal is really the core therapeutic ingredient of direct behavioral
change interventions. Within-session role plays are conducted in which the client practices executing the new behavior in a variety of possible situations. The therapist might
begin by modeling in the role play the target behavior such as initiating a conversation
with a stranger, making a request, maintaining eye contact, refusing an unreasonable
request, or the like. The client is then asked to practice the behavior within the role
play. Throughout the role play the therapist provides coaching in the form of corrective feedback as well as reinforcement and encouragement for attempts to perform the
target behavior. Since many individuals are uncomfortable with acting and may find
these behavioral practice sessions tedious, it is important to keep the atmosphere light
or informal and use humor to put individuals at ease. In the treatment of social phobia
videotaped in-session role plays with therapist and client or with additional “actors” can
be used to enhance behavioral rehearsal (e.g., Antony & Swinson, 2000a; D. M. Clark,
2001). In such cases the therapist provides feedback and correction while reviewing the
tape with the client.
Beck at al. (1985, 2005) also notes that important dysfunctional thoughts and
beliefs may become apparent in the course of behavioral rehearsal. Once identified these
automatic thoughts and beliefs would be addressed with cognitive restructuring strategies. For example, during behavioral rehearsal that targeted eye contact with a person
suffering from chronic social phobia, the therapist noticed that the client had great
258
ASSESSMENT AND INTERVENTION STRATEGIES
difficulty maintaining eye contact. The role play was stopped and the therapist asked
the client “When we were role-playing just now, what was going through your mind?”
The client stated that he was thinking “I am staring at the person; he is going to get
angry if I just keep staring like this.” So, automatically the client would break off his
gaze and look away, which meant that he did not perform the behavioral rehearsal correctly. Identifying and correcting faulty cognitions that arise in the course of behavioral
rehearsal is an important use of this strategy in cognitive therapy for anxiety.
The effectiveness of any behavioral change intervention will depend on whether
behavioral rehearsal is followed by systematic and repeated practice of these new skills
as in vivo homework assignments. As with any intervention the generalizability and
maintenance of any new learning achieved within session depends on completion of
homework assignments. Individuals should also self- monitor their behavioral homework assignments by keeping a record of the situations in which they practiced the new
behavior, their anxiety level, the outcome, and their evaluation of their performance. In
the follow-up session the therapist would review the homework self-monitoring form.
Examples of positive behavioral change would be praised and any problematic cognitions or behavioral responses would be targeted for further intervention.
Clinician Guideline 7.7
Direct behavioral change interventions are often employed in cognitive therapy to address
performance deficits in social functioning that may exacerbate withdrawal and isolation
from others and interfere with the client’s participation in crucial between-session exposure
assignments.
RELAXATION TRAINING
Relaxation training has had a long and venerable history in behavior therapy for anxiety. At one time it was the cornerstone of behavioral treatment for anxiety and considered critical for inhibiting conditioned anxiety responses (i.e., Wolpe & Lazarus, 1966).
Recently cognitive-behavior therapists have questioned the wisdom and effectiveness
of relaxation therapy for anxiety. White and Barlow (2002), for example, argued that
any behavior that minimizes panic symptoms or provides escape/distraction from these
symptoms would be maladaptive. Teaching individuals to relax via progressive muscle
relaxation or breathing retraining could undermine exposure and be tantamount to
“teaching avoidance as a coping strategy” (White & Barlow, 2002, p. 317). In many
respects relaxation training is also incompatible with the objectives of CT for anxiety. Empirical hypothesis testing of faulty appraisals and beliefs depends on exposure
to anxiety situations in order to gather disconfirming information. If relaxation was
invoked whenever a person felt anxious, then that person would forfeit an opportunity
to learn that the anxious concerns were unfounded. In this way relaxation as an anxiety
management response would undermine the effectiveness of cognitive therapy.
So, is there a place for relaxation training in cognitive therapy of anxiety? We would
only recommend relaxation techniques as an adjunctive intervention if an individual’s
Behavioral Interventions
259
anxiety level was so extreme that the client refused to engage in any exposure or refused
to tolerate even the slightest amount of anxiety. In such cases relaxation training could
be taught to lower anxiety level so the individual would engage in exposure and other
behavioral experiments designed to modify the faulty appraisals and beliefs of threat,
vulnerability, and the need for safety. For the cognitive therapist, it is the deactivation
of the fear schemas that is considered essential for long-lasting reduction in anxiety and
not the acquisition of a relaxation coping strategy.
Despite these concerns with its conceptual basis, relaxation training continues to
be advocated as an effective intervention for inhibiting the physical tension of anxiety (e.g., Bourne, 2000; Craske & Barlow, 2006). However, the empirical research
indicates that relaxation training has a far more limited role in treatment of anxiety
than once envisioned. Progressive muscle relaxation, for example, continues to be an
important therapeutic ingredient in CBT protocols for GAD (e.g., Brown, O’Leary, &
Barlow, 2001; see Conrad & Roth, 2007, for review of empirical status) and PTSD
(Foa & Rothbaum, 1998), but it appears to have less value for social anxiety (Heimberg & Juster, 1995) and OCD (Foa et al., 1998; Steketee, 1993), and has produced
mixed results, at best, for panic disorder (see D. M. Clark, 1997; Craske & Barlow,
2001, for reviews).
Progressive Muscle Relaxation
In 1938 Edmund Jacobson published his work on relaxation that was based on a rather
unique theory of anxiety. Jacobson argued that the core experience of anxiety is muscle
tension, which involves contraction or shortening of the muscle fibers. In order to reduce
this tension and subjective anxiety, progressive muscular relaxation (PMR) was introduced as a method that eliminates tension by lengthening muscle fibers (Jacobson, 1968;
see also Bernstein & Borkovec, 1973). By systematically tensing and releasing various
muscle groups, Jacobson found that muscle contractions could be practically eliminated
and a state of deep relaxation induced. The only problem is his method of relaxation
was extremely time consuming, involving 50–200 sessions of training (see Wolpe, 1958;
Wolpe & Lazarus, 1966).
Jacobson’s relaxation procedure was adopted and refined by the pioneers of
behavior therapy as an incompatible response that could inhibit fear and anxiety.
Wolpe (1958) concluded from Jacobson’s writings that his relaxation method had
anxiety- countering effects, because individuals were taught to use differential relaxation in their day-to-day lives in which muscle groups not directly in use were relaxed.
This will lead to reciprocal inhibition of any anxiety- evoking stimuli encountered and
with repeated occurrences a conditioned inhibition of the anxiety response gradually
develops. However, Wolpe (1958) introduced two major modifications to improve the
efficiency and effectiveness of differential relaxation. First, he was able to drastically
reduce the number of relaxation training sessions to six 20-minute sessions and two
15-minute daily practice sessions at home (Wolpe & Lazarus, 1966). And second, in
subsequent sessions relaxation was paired with systematic graduated imaginal evocation of a fear stimulus in a treatment procedure called systematic desensitization. The
result was the introduction of a highly effective behavioral treatment for fears and
phobias.
260
ASSESSMENT AND INTERVENTION STRATEGIES
The induction of deep relaxation became an essential tool in the behavior therapist’s
armamentarium for inhibiting anxiety. Wolpe discovered that the autonomic effects of
relaxation can only counter a weak anxiety response, but once a weak stimulus is no
longer anxiety-provoking a slightly stronger anxiety-provoking stimulus can be repeatedly paired with relaxation until it too ceases to arouse anxiety (Wolpe & Lazarus,
1966). Gradually, with repeated presentations, deep relaxation will inhibit successively
stronger anxiety responses until even the most intense anxiety-provoking situation no
longer elicits anxiety.
The systematic tensing and releasing of specific muscle groups that was pioneered
by Edmund Jacobson is still the most common approach to relaxation training used in
CBT. Clients are instructed to tense a specific muscle group “as hard as possible without
causing pain,” to hold the tension for 5–7 seconds, to notice the tension in the muscle
group, then to relax and release the tension, and to notice the feeling of relaxation that
occurs when the tension is released (Bernstein & Borkovec, 1973). The purpose of this
“tense–release” cycle is to facilitate tension detection and sharpen the client’s ability to
discriminate between sensations of tension and relaxation. Although many different
PMR variations exist, we present a 10-muscle group protocol in Table 7.5 that can be
initially taught to clients. It is derived from lengthier protocols described in Bernstein
and Borkovec (1973), and Cautela and Groden (1978).
Rationale and Instructions
Before initiating a relaxation training session, it is important to provide a rationale for
the procedure. The following is one possible explanation and set of instructions for
PMR that can be used with clients. (For other examples of rationales and instructions
for PMR, see Bernstein & Borkovec, 1973; Bourne, 2000; Cautela & Groden, 1978;
Craske & Barlow, 2006; Foa & Rothbaum, 1998; Goldfried & Davison, 1976.)
“Today I am going to teach you how to use relaxation to mange your anxiety.
This procedure, called deep muscle relaxation, was first introduced 75 years ago
by a Harvard University physiologist, Dr. Edmund Jacobson. He found that individuals could learn to induce a state of deep relaxation by tensing and then releasing specific groups of muscles. The important part of this procedure is learning the
difference between feeling tense and feeling relaxed, so you will be coached on how
to pay especially close attention to the feelings and physical sensations associated
with your muscles being tense and then relaxed. Do you recall from the earlier
assessment session that one of the anxiety symptoms that you noticed was muscle
tension? Could you remind me of what that is like for you. [Have client describe
the discomfort associated with feeling physically tense or uptight when anxious.]
When you feel tense, certain muscles in your body tighten; that is, the muscle fibers
actually contract, producing that tense feeling. Progressive muscle relaxation is
a technique that interrupts the anxiety process by relaxing the muscles. It literally reverses one of the main symptoms of anxiety, physical tension, by releasing
unwanted muscle contraction or tension. Once you’ve mastered the skill of inducing
deep relaxation, you can use it in a variety of situations to interrupt a rise in your
anxiety level.
Behavioral Interventions
261
TABLE 7.5. A 10-Muscle-Group Protocol for Progressive Muscle Relaxation
Muscle group
Tense–release procedure
1. Dominant arm
“Extend your right arm (i.e., dominant) straight out, make a tight fist, and
tighten whole arm from hand to shoulder. Notice tension in biceps, forearm,
elbow, wrist, and fingers. Then relax, bending arm at elbow and resting it on
your lap.”
2. Nondominant arm
“Extend your left arm (i.e., nondominant) straight out, make a tight fist, and
tighten whole arm from hand to shoulder. Notice tension in biceps, forearm,
elbow, wrist, and fingers. Then relax, bending arm at elbow and resting it on
your lap.”
3. Forehead
“Wrinkle forehead by lifting the eyebrows as high as you can, push your
eyebrows up, putting tension in the forehead and scalp areas. Then slowly relax,
letting your eyebrows drop and notice the release of tension in the forehead.”
4. Eyes and nose
“Close your eyes very tightly, squint them hard so you can feel tension around
your eyes. At the same time, wrinkle your nose, again pushing your nose hard
against your face. Notice the tension around the eyes, nose, and upper cheeks.
Slowly relax, release the tension around your eyes and nose by not squinting
your eyes or wrinkling your nose. Keep your eyes closed and focus on the
relaxed feelings around your eyes and nose.”
5. Jaw and neck
“Tense the mouth, jaw, and neck regions by making an exaggerated grin, clench
the teeth, and tighten your neck by drawing your mouth and chin inward.
Notice the tightness of your muscles around the mouth, jaw, and front part of
the neck. As you release the tension, focus on the feeling of relaxation in these
regions of the face and neck.”
6. Shoulders and back
“Move forward in the chair and bring the elbows up and back so that you can
feel your shoulder blades being pushed together. At the same time the chest is
being pulled out. Notice the tension in the shoulders and upper back. Gradually
release the tension by sitting back in the chair, placing your arms in your lap
and allowing the shoulders to fall back into their normal position. Focus on the
release of tension in the shoulders and down the middle of the back.”
7. Chest
“Tighten the chest by taking a deep breath and then hold it. Feel the tension in
the chest as you constrict and pull it in. As you relax focus on how loose the
chest muscles now feel.”
8. Stomach
“Tighten the stomach by pulling it in and making it as hard as a board. Notice
the tension in your stomach and how hard it feels. As you release the tightness
in your stomach, notice how it feels to switch from tension to relaxation.”
9. Dominant leg
“Lift your right (i.e., dominant) leg off the floor so that your leg is fully
extended outward, bend your toes inward toward you, and tighten your whole
leg as much as possible. Notice the tension in your foot, calf, knees, and
thighs. Gradually relax, lowering your leg back to the floor and bending your
knee slightly so that your foot is squarely on the floor. Notice the feeling of
relaxation that now permeates through the entire length of the leg.”
10. Nondominant leg
“Lift your left (i.e., nondominant) leg off the floor so that your leg is fully
extended outward, bend your toes inward toward you, and tighten your whole
leg as much as possible. Notice the tension in your foot, calf, knees, and
thighs. Gradually relax, lowering your leg back to the floor and bending your
knee slightly so that your foot is squarely on the floor. Notice the feeling of
relaxation that now permeates through the entire length of the leg.”
Note. Based on Bernstein and Borkovec (1973) and Cautela and Groden (1978).
262
ASSESSMENT AND INTERVENTION STRATEGIES
“The best way to learn deep muscle relaxation is through demonstration,
coaching, and practice. I am going to ask you to produce tension in particular
muscle groups, hold the tension for 5–7 seconds, and then release this tension. I will
instruct you on how to tense and release various muscles. Throughout the procedure I will be prompting you to focus on the feelings of tension and relaxation. This
is a very important part of the technique because you need to learn how it feels to be
relaxed. We will begin by tensing and relaxing 10 different muscle groups and the
whole procedure will take about 20 minutes. I will be asking you to tense and then
relax particular muscles. For example, let’s quickly run through the procedure with
each of the muscles so you will know what to expect. Take your right arm, extend
it in front of you, make a tight fist, and hold it. Do you notice any tension or tightness in your arm? [Ask client to indicate whether tension was felt in hands, forearm,
elbow, and bicep of the arm.] Now tense the arm again and this time release the
tension by letting your arm fall back into your lap, with the arm slightly bent at the
elbow. How does it feel now? [Client is asked to describe the feeling of relaxation in
the arm.] Now I am going to demonstrate for you how to tense and relax the other
9 muscle groups. Each time I would like you to watch how I do it and then try for
yourself. I do need to warn you we’ll be making some funny faces in order to tense
the facial muscles. Are you okay with that? [Therapist then demonstrates how to
tense and relax muscles based on Table 7.5.]
“[After demonstrating the 10-muscle tense–release procedure, the therapist
continues with the introduction.] It is important that you realize that deep muscle
relaxation is a skill that takes repeated practice to learn. Just like learning to ride
a bicycle or drive a car, the technique may at first feel unnatural to you. You may
not feel very relaxed. However, the more you practice it, the easier it will become
and you will get better and better at inducing a deeper level of relaxation. Also once
you’ve mastered the 10 muscle technique, I will teach you how to do the abbreviated version of muscle relaxation so that you can literally induce relaxation in a
few minutes anywhere, anytime. But to get to that point, you will need to practice
relaxation twice a day, every day for 15 minutes. I will be giving you a CD with
relaxation instructions that should help you do the homework practice. Also I will
ask that you complete a Weekly Progressive Muscle Relaxation Record [see Appendix 7.4] so we can monitor your progress. Do you have any questions? Okay, let’s
begin with our first relaxation training session.”
It is important to emphasize that the effectiveness of relaxation training depends on
a conducive setting. Bourne (2000) offers a number of practical suggestions for enhancing the relaxation experience. Choose a quiet location, a dimly lit room, and a comfortable chair or sofa. Practice on an empty stomach and loosen any tight-fitting garments.
Remove shoes, watches, and glasses, and keep eyes closed. Tell the client to assume
a passive, detached attitude in which “you let everything, all thoughts, feelings, and
behavior, just happen. Don’t try to control what you are thinking or evaluate how you
are performing. Just ‘let yourself go’ and don’t worry about whether you are doing the
procedure correctly.” If the person has difficulty relaxing a particular muscle group,
he or she should just skip to the next group of muscles. Not all muscle groups have to
achieve the same level of deep relaxation. Emphasize that it is important to practice
twice a day for 15 minutes preferably at a regular time.
Behavioral Interventions
263
The following example illustrates how to coach a client on the tense–release cycle.
We have chosen the stomach muscle group to illustrate the instructional set that should
be employed with each muscle group.
“Now I would like you to tense your stomach muscles. Tighten your stomach by
pulling it in and making is as hard as a board. NOW, tighten your stomach muscles
[therapist uses firm, moderately loud voice]. HOLD IT! Feel the tension, the tightness
of your stomach muscles, HOLD IT, HOLD IT! Focus your attention on the hardness of your stomach [5–7 seconds after the NOW] And noowww, RELAXXXX!
[Therapist drags out the “now relax” in a lower, soothing voice.] Let all the tension
go from your stomach, let it flow out of your muscles, and notice the difference
between feeling tense and relaxed. You feel your stomach muscles go further and
further into relaxation. [For 30–40 seconds therapist makes suggestive statements
about relaxation.] You focus all your attention on the pleasant feeling of relaxation.
You notice how the stomach muscles now feel slack, loosened, and smoothed out
compared to their hard, stiff, and tight state when you were tensing them. Continue
to focus your attention on the feeling of relaxation as we move to your right leg.”
In the first training session of PMR, it may be advisable to repeat each muscle group
twice before proceeding to the next set of muscles. Also allow a few seconds of silence
between muscle groups so that the whole process does not become too hurried. During
each phase of release the client should subvocally repeat the word “relax” or “calm.”
Furthermore, the therapist can add a pleasant imagery suggestion at the end of the
relaxation session in order to enhance the experience of deep relaxation.
Abbreviated PMR
If PMR is to have any utility as a coping response for anxiety in the naturalistic setting,
clients must quickly learn more efficient, abbreviated relaxation protocols that can be
employed anytime and in any place. If the client has mastered 10-muscle deep relaxation after 2 weeks of daily practice, the therapist can proceed with a 4-muscle group
protocol described in Bernstein and Borkovec (1973). This consists of the following
procedure:
1. Tense and release the arms—both arms are held out in front of the person with
a 45° bend at the elbow. Make a tight fist in each hand and hold the tension.
2. Face and neck—all of the face and neck muscles are tensed simultaneously by
making a frown, squinting the eyes, wrinkling the nose, clenching the teeth,
making an exaggerated grin, and pulling the chin downward to the chest.
3. Chest and abdomen—take a deep breath and then hold it while at the same time
sitting forward, pull the shoulders back so that the shoulder blades are being
pushed together, and tighten the stomach.
4. Both legs—lift both legs off the floor, point the toes up, and rotate the feet
inward.
If deep muscular relaxation can be achieved after 2 weeks of daily practice, the client is ready to proceed to the final stage of PMR, release-only relaxation. Here the tense
264
ASSESSMENT AND INTERVENTION STRATEGIES
part of the exercise is omitted and the client simply focuses on releasing tension in various muscle groups starting at the top of the head and progressing downward to the toes
(Taylor, 2000). Having daily practiced deep muscle relaxation for at least a month, individuals are now so well accustomed to the relaxed state that they are able to feel relaxed
simply through recall (Bernstein & Borkovec, 1973). When asked to release the tension
from particular muscle groups, this can be done by recalling their previous relaxed
state. In release-only relaxation, the client is first instructed to breathe calmly and then
to relax the various muscles of the face, neck, shoulders, arms, stomach, back, and
legs (see Öst, 1987a, for detailed instructions). Once again individuals should practice
release-only relaxation twice per day for at least 1 week. The protocol can be recorded
to assist with homework practice and then faded out as the client masters this skill (Taylor, 2000). Clients who have mastered release-only relaxation now have a coping skill
that can be used in almost any situation involving naturally occurring anxiety. It is a
highly portable, efficient technique that enables the individual to achieve a relaxed state
in 5–7 minutes (Öst, 1987a).
Clinician Guideline 7.8
Progressive muscle relaxation is an adjunct intervention that can be used by the cognitive
therapist as preliminary skills training to reduce extreme levels of anxiety so the client will
engage in self-directed exposure or to provide coping strategies for individuals with severe
intolerance of anxiety. However, any relaxation training must be carefully monitored to
ensure it is not used to avoid anxiety or to undermine the benefits of exposure-based behavioral experimentation.
Applied Relaxation
Applied relaxation (AR) is an 8- to 10-week treatment program developed by LarsGöran Öst (1987a) at the Psychiatric Research Center, University of Uppsala, Sweden.
It is an intensive, systematic, graded form of relaxation training that builds from PMR
through cue- controlled relaxation to the application of rapid relaxation skills to anxiety elicited in natural situations. Because the final stage of AR involves within- and
between-session practice in applying relaxation to anxiety-arousing situations, AR
actually involves repeated brief situational and interoceptive exposure and so can not be
considered a purely relaxation-based intervention for anxiety (Taylor, 2000). Nevertheless, what makes AR of interest is its conceptualization in terms of a coping perspective
on anxiety and empirical evidence of its effectiveness for GAD in particular (e.g., see
meta-analysis by Gould, Safren, Washington, & Otto, 2004). Öst (1987a) states that the
purpose of AR is to teach individuals how to recognize the early signs of anxiety and
learn to cope with anxiety rather than feel overwhelmed by their anxiousness. Table 7.6
presents a breakdown of the AR procedure as described by Öst (1987a).
Öst (1987a) reviewed 18 controlled outcome studies from his own lab that utilized AR and concluded that 90–95% of individuals were able to acquire the relaxation
skill, with AR significantly more effective than no treatment or nonspecific treatment
comparisons. The strongest empirical evidence for the effectiveness of AR comes from
Behavioral Interventions
265
TABLE 7.6. Applied Relaxation Treatment Protocol
Sessions
Intervention
Instructions
Session 1
Psychoeducation
Explain nature of anxiety, rationale for AR, graduated
homework in identifying and recording symptoms of
anxiety.
Sessions 1–4
14-muscle PMR
Complete body relaxation based on the 14-muscle
PMR protocol of Wolpe & Lazarus (1966). Twice daily
homework practice assigned.
Sessions 5–6
Release-only relaxation
Teach relaxation of muscle groups directly without
tension instructions. Reduce relaxation induction time to
5–7 minutes. Takes 1 or 2 sessions with daily homework
practice.
Sessions 6–7
Cue-controlled relaxation
Purpose is to create conditioned association between word
“relax” and the relaxation state. Focus is on controlled
breathing, relaxation induced via release-only method, and
repeated pairing of subvocalization of word “relax” with
every exhale. Homework practice assigned for 1–2 weeks.
Sessions 8–9
Differential relaxation
Purpose is to teach individuals to relax in other situations
such as seating at a desk or walking and to remove tension
from muscles not in use for an activity.
Sessions 10
Rapid relaxation
Teach client to relax in 20–30 seconds in multiple
nonstressful daily situations by controlled breathing,
think “relax,” and scan body for tension and release by
relaxation.
Sessions 11–13
Application training
Brief exposure (10–15 minutes) to wide range of in vivo
anxiety-arousing situations, physical sensations (i.e.,
hyperventilation, physical exercise), or imagery in order to
practice applying relaxation as coping response to anxiety.
Sessions 14–15
Maintenance program
Client encouraged to scan body at least daily and use rapid
relaxation to get rid of any tension. Differential and rapid
relaxation to be practiced twice a week on a regular basis.
clinical trials of GAD. In a variety of outcome studies AR produced significant posttreatment effects for GAD and maintenance of gains over follow-up that equalled cognitive therapy (Arntz, 2003; Borkovec & Costello, 1993; Borkovec, Newman, Lytle, &
Pincus, 2002; Öst & Breitholz, 2000). However, Butler, Fennell, Robson, and Gelder
(1991) found that standard PMR was less effective than cognitive therapy for GAD and
barely more effective than a wait-list control. Moreover D. M. Clark and colleagues
found that cognitive therapy was somewhat superior to AR in the treatment of panic
disorder (D. M. Clark et al., 1994) and clearly superior to AR plus exposure in the
treatment of social anxiety (D. M. Clark, Ehlers, Hackmann, McManus, Fennell et
al., 2006). Öst and Westling (1995), on the other hand, found that CBT and AR were
equally effective in treatment of panic disorder. In summary, it would appear that AR
is an alternative treatment for GAD that can produce results equivalent to cognitive
therapy, but its effectiveness for the other anxiety disorders remains less certain.
266
ASSESSMENT AND INTERVENTION STRATEGIES
Clinician Guideline 7.9
Applied relaxation (AR) is an intensive, systematic, and graded relaxation training protocol
that can be effective in the treatment of GAD, although it may be less effective for other anxiety disorders. AR is a viable alternative to cognitive therapy for GAD when the latter may not
be acceptable to a client.
Breathing Retraining
Training in controlled breathing is considered a form of relaxation that is often included
in relaxation procedures for stress and anxiety (e.g., Bourne, 2000; Cautela & Groden,
1978). Individuals often engage in rapid shallow breathing when in anxious or stressful
situations. Controlled breathing procedures train individuals to become more aware of
their dysfunctional breathing and to replace this with a slower, more paced diaphragmatic breathing of approximately 8–12 breaths per minute. This slower, deeper rate of
breathing promotes a greater sense of relaxation, thereby reducing the anxious state. It
is a quick and fairly simple intervention strategy that can give anxious individuals a limited sense of control over their emotional state. Because breathing retraining has been
used most extensively in CBT of panic disorder, further discussion of this procedure is
presented in the next chapter.
Clinician Guideline 7.10
Controlled breathing is a relatively quick and simple relaxation strategy that can be used to
counter the rapid, shallow overbreathing that often contributes to heightened anxiety. In
recent years clinical research has questioned the therapeutic role of controlled breathing,
particularly in the treatment of panic disorder.
SUMMARY AND CONCLUSION
Behavioral interventions play a critical role in cognitive therapy of anxiety disorders. In
fact it is difficult to imagine an effective cognitive treatment for anxiety that does not
include a significant behavioral component. There is a large empirical literature demonstrating the effectiveness of exposure interventions in the treatment of all types of fear
and anxiety. When utilized as a therapeutic ingredient of cognitive therapy, exposurebased exercises provide the most powerful forms of corrective information for the faulty
threat and vulnerability appraisals and beliefs that sustain heightened anxiety. Exposure in the form of empirical hypothesis-testing experiments should be a focal point in
all cognitive therapy interventions offered to treat the anxiety disorders.
Greater attention should be given to response prevention and correction of safetyseeking cognitions and behaviors in cognitive interventions for anxiety (e.g., D. M. Clark
et al., 1999; Salkovskis, Clark, & Gelder, 1996). Without intervention that directly
reduces reliance on safety-seeking cues and coping responses, it is likely that any reduction in anxiety will be incomplete and place the individual at high risk for relapse.
Behavioral Interventions
267
The role of relaxation training in treatment of anxiety disorders continues to generate considerable debate. The long-established tradition of teaching progressive muscle
relaxation to relieve anxiety may still have some efficacy for the treatment of GAD and
possibly panic disorder, especially when the more systematic and intense applied relaxation protocol is employed. However, relaxation training for OCD and social phobia is
unwarranted, although it may still have some value in PTSD for those with heightened
generalized anxiety. Breathing retraining is often used in treatment of panic disorder
but as discussed in the next chapter its therapeutic effectiveness has been called into
question. A Quick Reference Summary is provided in Appendix 7.5 as a brief overview
of the behavioral interventions that are useful in treating the anxiety disorders.
APPENDIX 7.1
Exposure Hierarchy
Name:
Date:
Instructions: On a blank sheet of paper write down 15–20 situations, objects, physical sensations,
or intrusive thoughts/images that are relevant to your anxious concerns. Select experiences that fall
along the full range from those that trigger only slight anxiety and avoidance to experiences that
elicit moderate and then severe anxiety and avoidance. Next rank-order these experiences from
least to most anxious or avoidant and transfer the list into the second column on this form. In the
first column record the level of anxiety you expect with each entry. In the third column write down
the core anxious thought associated with each situation if this is known to you.
A. Expected
level of anxiety/
avoidance (0–100)
LEAST
B. Briefly describe the anxious/
avoided situation, object, sensation,
or intrusive thought/image
C. Note the most prominent
anxious or apprehensive
thought associated with
this entry
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
MOST
17.
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
268
APPENDIX 7.2
Exposure Practice Record
Name:
Date:
Instructions: Keep a record of your daily exposure practice sessions using this form. Be sure to
record the initial, middle, and final anxiety rating as well as the type of exposure task completed
and its duration.
Date
and
Time
EndInitial Midpoint point
Duration Anxiety Anxiety Anxiety
(minutes) (0–100) (0–100) (0–100)
Exposure Task
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
269
APPENDIX 7.3
Response Prevention Record
Name:
Date:
Instructions: Keep a record of your daily response prevention practice sessions using this form. Be
sure to record the initial and endpoint “urge to engage in response” and anxiety level.
Date
and
Time
Describe Response That Was Prevented
Initial
Endpoint
Urge to
Urge to Endpoint
Engage in Initial Engage in Anxiety
Response Anxiety Response
Level
(0–100) (0–100) (0–100) (0–100)
List “blocking strategies” used for response prevention:
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
270
APPENDIX 7.4
Weekly Progressive Muscle Relaxation Record
Name:
Date:
Instructions: Two 15-minute relaxation sessions should be scheduled daily. Use the chart below to record your progress in achieving a relaxed
state with each of the muscle groups. Make a check mark () if you successfully relaxed a muscle group during a practice session and mark
an (X) if you had difficulty relaxing the muscle group. At the bottom of the column, rate the overall level of relaxation achieved in the practice
session from 0 (“unable to relax at all”) to 50 (“moderately relaxed but conscious of some tension”) to 100 (“so completely relaxed that I fell
asleep”).
Day of Week:
Day One
Day Two
Day Three
Day Four
Day Five
Day Six
Practice session:
1
1
1
1
1
1
2
2
2
2
2
2
Day Seven
1
2
1. Dominant arm
2. Nondominant arm
271
3. Forehead
4. Eyes and nose
5. Jaw and neck
6. Shoulders and back
7. Chest
8. Stomach
9. Dominant leg
10. Nondominant leg
11. Rate overall level of relaxation
(0–100)
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright 2010 by The Guilford Press. Permission to photocopy
this appendix is granted to purchasers of this book for personal use only (see copyright page for details).
APPENDIX 7.5
Chapter 7 Quick Reference Summary: Behavioral Interventions
I.
Adopt a Cognitive Perspective
1. Rationale—based on Figure 6.1 (client handout of cognitive therapy model), explain use of behavioral
assignment to examine validity of anxious thoughts and their alternatives.
2. Identify Target Thought—write down the anxious thought challenged by the behavioral exercise.
3. Behavioral Prescription—write out specific instructions on how to do exercise, what thoughts are
evaluated, and the outcome criteria.
4. Self-Monitoring—client records how the exercise was conducted, its outcome, anxiety level,
automatic thoughts, evidence for and against target thoughts.
5. Evaluation—extensive evaluation of outcome of exercise; review self-monitoring form; conclusions
reached about target thought (belief) and its alternative; write out a summary of exercise in form of a
“coping card.”
II.
Graded Exposure
1. For situational exposure, review the Situational Analysis Form (Appendix 5.2) and hierarchically
arrange anxiety-provoking situations from mildly to intensely anxious.
2. Begin with moderately anxious situation; initially demonstrate exposure within session.
3. Obtain 0–100 anxiety ratings before exposure, every 10 minutes during exposure, and finally at
conclusion of the exercise.
4. Assign exposure as homework, at least 30–60 minutes daily. Use Exposure Practice Record
(Appendix 7.2) to record outcome.
5. Imaginal exposure begins with development of a fear script, within-session demonstration, and then
30 minutes of daily homework. Audio habituation training should be considered when cognitive
avoidance is present.
6. Exposure to bodily sensations involves extensive within-session demonstration prior to homework
assignment. Table 8.8 (panic disorder chapter) provides a description of various interoceptive
exercises.
III. Response Prevention
1. Identify maladaptive cognitive and behavioral coping strategies or other forms of neutralization
(see Behavioral Responses to Anxiety Checklist, Appendix 5.7, and Cognitive Responses to Anxiety
Checklist, Appendix 5.9).
2. Provide treatment rationale for response prevention.
3. Instruct client on “blocking strategies” (e.g., self-instructional coping statements, competing
responses, paradoxical intention, encouragement).
4. Develop alternative coping strategies for anxiety.
5. Challenge problematic cognitions.
6. Record and evaluate success of intervention using the Response Prevention Record (Appendix 7.3).
IV. Other Behavioral Interventions
1. Direct behavioral change involves teaching specific behaviors that improve personal effectiveness
through methods of didactic instruction, modeling, behavioral rehearsal, reinforcement, and selfmonitoring.
2. Relaxation training can be progressive muscle or applied relaxation training; most useful for GAD. A
rationale for PMR can be found in Chapter 7, pages 260–262. Instructions for 10-muscle PMR are in
Table 7.5 and an outline for AR is described in Table 7.6. Assign PMR as homework and record daily
practice on the Weekly Progressive Muscle Relaxation Record (Appendix 7.4).
3. Breathing retraining—Table 8.9 on page 324 (panic disorder chapter) contains diaphragmatic
breathing retraining protocol.
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
272
PART III
COGNITIVE THEORY AND TREATMENT
OF SPECIFIC ANXIETY DISORDERS
In the last two decades psychotherapy innovation and research has focused
increasingly on the development and evaluation of treatment protocols that
target specific DSM-IV-TR (APA, 2000) disorders. The growth of disorderspecific manualized treatment has been particularly evident in the anxiety
disorders. The generic cognitive model of anxiety presented in Part I and the
core cognitive assessment and intervention strategies described in Part II can
be readily adapted to target the shared and distinct symptom features of the
more common types of anxiety disorders. This final part of the book provides
disorder-specific cognitive models, hypotheses, case conceptualizations, and
treatment protocols for five different types of anxiety disorder. Chapter 8 discusses the cognitive model and treatment of panic disorder with its emphasis
on threat misinterpretations of internal states and loss of reappraisal capacity,
whereas Chapter 9 presents the cognitive theory and treatment of social phobia
that focuses on fear of negative evaluation of others and presence of maladaptive coping responses. Chapter 10 provides a cognitive model and treatment of
generalized anxiety and worry, Chapter 11 discusses the cognitive appraisal
perspective on theory and treatment of obsessive– compulsive disorders, and
Chapter 12 presents a cognitive model and treatment that focuses on the faulty
appraisals and beliefs associated with the trauma-related intrusive thoughts
and memories of posttraumatic stress disorder.
273
Chapter 8
Cognitive Therapy of Panic Disorder
Of course, we were afraid and fear isn’t always a wise counselor,
let’s go back, for our greater safety we ought to barricade the
door of the wards. . . .
—JOSÉ SARAMAGO (Portuguese novelist and 1998
Nobel Laureate in Literature, 1922– )
Helen is a 27-year-old single woman who worked in the insurance industry and
presented with an 11-year history of panic disorder and moderate agoraphobic
avoidance. At the time of assessment she was experiencing approximately eight
full-blown panic attacks daily with elevated levels of generalized anxiety, considerable apprehension about having panic attacks, and avoidance of routine
activities such as travel outside her community, not maintaining close proximity to medical facilities, highway driving, air travel, and the like. The first
onset of panic occurred when she was 16 years old but the panic attacks were
few and far between until she took her first business trip to New York City at
age 22. She described 4 days of terrifying acute anxiety involving chest pain,
heart palpitations, tingling in the extremities, abdominal distress, and agitation. These bodily sensations were accompanied by an intense fear that she
might die from a heart attack. However, she did not seek medical intervention
at the time but instead coped by resting, taking Gravol, and trying to remain
calm. Upon returning home the panic attacks continued. In the intervening 5
years she has been treated with citalopram, lorazepam, and relaxation training
with minimal effectiveness.
Pretreatment assessment revealed that heart palpitations, chest pain,
sweating, shortness of breath, feelings of choking, nausea, and hot flushes
were the main bodily sensations during her panic attacks. Although fears of a
heart attack or of going crazy were still present, her main misinterpretation of
threat had shifted to a focus on breathlessness, with a fear that she would stop
breathing and suffocate. Extensive reliance on safety seeking emerged such
that Helen became preoccupied with maintaining close geographic proximity
to medical facilities, frequently making trips to her family physician and hospital emergency department whenever she felt intense panic or concern about her
275
276
TREATMENT OF SPECIFIC ANXIETY DISORDERS
respiration or cardiac functioning. As a result she became increasingly reluctant to venture more than a few miles from a hospital for fear that she would
be trapped without access to medical facilities. Avoidance, reassurance seeking and self-monitoring of physical symptoms (e.g., repeated pulse checking)
became the main coping strategies for her daily battle with panic attacks.
A structured diagnostic interview revealed that Helen met DSM-IV criteria for panic disorder with agoraphobic avoidance of moderate severity. She
did not have any other current comorbid condition but did report two previous
episodes of major depression with suicidal ideation. Her pretreatment symptom scores were BDI-II = 8, BAI = 22, PSWQ = 64, Agoraphobic Cognitions
Questionnaire (ACQ) = 33, and Body Sensations Questionnaire (BSQ) = 48.
Her main threat-related thoughts concerned “What if I can’t get my breath and
I suffocate?”, “Could this chest pain mean that I am having a heart attack?”,
“What if I can’t get to the hospital in time?”, “What if this builds into another
panic attack and it eventually drives me crazy?”, and “Is this ever going to
end?” In short, Helen revealed a pattern of anxious thinking and misinterpretation that reflected an intolerance of anxiety and reliance on maladaptive
avoidance and safety-seeking strategies in a desperate attempt to control her
anxiety and prevent the much dreaded panic attacks.
Helen’s clinical state exemplifies a fairly typical presentation of panic disorder.
Twelve individual sessions of CBT followed by four booster sessions over an 8-month
period proved highly effective in reducing panic frequency, generalized anxiety, and
agoraphobic avoidance. Treatment focused on (1) psychoeducation in the cognitive therapy model, (2) intentional activation of bodily sensations and underlying fear schemas,
(3) cognitive restructuring and reattribution of misinterpretations of bodily sensations,
(4) graded situational exposure homework, and (5) increased tolerance and acceptance
of anxiety, risk, and uncertainty with a corresponding reduction in intentional control
efforts. In this chapter we begin with a description of the phenomenology and diagnosis
of panic and agoraphobia, followed by a discussion of the cognitive model of panic, and
its empirical status. The remainder of the chapter discusses issues of assessment, case
formulation, the cognitive therapy treatment protocol, and its efficacy.
DIAGNOSTIC CONSIDERATIONS AND CLINICAL FEATURES
The Nature of Panic
Panic attacks are discrete occurrences of intense fear or discomfort of sudden onset
that are accompanied by a surge of physiological hyperarousal. Barlow (2002) considers panic the clearest clinical presentation of fear. In addition to strong autonomic
arousal, panic is characterized by a faulty verbal or imaginal ideation of physical or
mental catastrophe (e.g., dying, going insane), intense uncontrollable anxiety, and a
strong urge to escape (Barlow, 2002; Beck et al., 1985, 2005; Ottaviani & Beck, 1987).
So aversive is the panic experience that many patients have a strong apprehension about
having another attack and develop extensive avoidance of situations thought to trigger
panic. As a result panic and agoraphobia are closely associated, with most individuals
with panic disorder presenting with some degree of agoraphobic avoidance and 95% of
Panic Disorder
277
people with agoraphobia reporting a past or current panic disorder (Antony & Swinson,
2000a; APA, 2000). In the latest epidemiological study panic disorder had a 12-month
prevalence of 2.7%, whereas agoraphobia without panic disorder was much less common at 0.8% (Kessler et al., 2005).
DSM-IV-TR defines panic attacks as “a discrete period of intense fear or discomfort in which four (or more) of the following symptoms developed abruptly and reached
a peak within 10 minutes” (APA, 2000, p. 432). The typical panic attack lasts between
5 and 20 minutes, although a heightened state of anxiety can linger long after the panic
episode subsides (Rachman, 2004). According to DSM-IV-TR, the defining symptoms
of panic are:
•฀ Elevated heart rate or palpitations
•฀ Sweating
•฀ Trembling or shaking
•฀ Smothering sensation or shortness of breath
•฀ Feeling of choking
•฀ Chest tightness, pain, or discomfort
•฀ Abdominal distress or nausea
•฀ Dizziness, lightheadedness, faintness, or feeling unsteady
•฀ Feelings of unreality (derealization) or detachment from oneself (depersonalization)
•฀ Numbness or tingling sensations
•฀ Chills or hot flushes
•฀ Fear of losing control or going crazy
•฀ Fear of dying
Table 8.1 lists a number of prominent features that characterize panic attacks.
Situational Triggers
Even though DSM-IV specifies that two unexpected panic attacks must occur to meet
diagnostic criteria for panic disorder, the majority of panic episodes are anticipated
because they are provoked by exposure to an identifiable stressor (Rachman, 2004).
Theaters, supermarkets, restaurants, department stores, buses, trains, airplanes, subTABLE 8.1. Critical Features of Panic Attacks
•฀ Situational triggers
•฀ Abrupt onset of physiological arousal
•฀ Heightened self-focus, hypervigilance of bodily sensations
•฀ Perceived physical, mental, or behavioral catastrophe
•฀ Apprehension, fear of future panic attacks
•฀ Extensive safety seeking (escape, avoidance, etc.)
•฀ Perceived lack of controllability
•฀ Qualitatively distinct from anxiety
278
TREATMENT OF SPECIFIC ANXIETY DISORDERS
ways, driving in the car, walking on the street, staying alone at home, or being far away
from home are all examples of external situations that individuals with panic disorder
report may trigger a panic attack. As a result these situations are often avoided in order
to minimize the possibility of triggering a panic episode. More recently, researchers have
argued that internal cues such as thoughts, images, feelings, or bodily sensations can
trigger panic and avoidance (Barlow, 2002; McNally, 1994; White et al., 2006).
Acute Physiological Arousal
Although an abrupt onset of physiological symptoms is one of the hallmarks of panic
attacks, it is clearly not a defining feature of the disorder. Individuals with panic disorder are not more autonomically hyperactive to standard laboratory stressors than
nonpanickers (Taylor, 2000). Furthermore, even though 24-hour ambulatory heart rate
monitoring of panic patients indicates that most panic attacks involve a distinct elevation in heart rate, a significant minority of self-reported attacks (i.e., 40%) are not associated with actual increase in heart rate or other physiological responses and most episodes of physiological hyperarousal (i.e., tachycardia) occur without self-reported panic
episodes (e.g., Barsky, Cleary, Sarnie, & Rushkin, 1994; Lint, Taylor, Fried-Behar, &
Kenardy, 1995; Taylor et al., 1986). Moreover, individuals with panic disorder do not
have more cardiac arrhythmias in a 24-hour period than nonpanic patients investigated
for heart palpitations (Barsky et al., 1994). As discussed below, it is not the presence of
physiological symptoms that is critical in the pathogenesis of panic but rather how these
symptoms are interpreted.
Hypervigilance of Bodily Sensations
Empirical studies are inconsistent on whether panic disorder is characterized by heightened interoceptive acuity especially in terms of cardiac perception (e.g., Pollock, Carter,
Amir, & Marks, 2006), although individuals may be more sensitive to the particular
body sensations linked to their central fear (e.g., increased pulse rate for those afraid of
heart attacks; Taylor, 2000). As McNally (1999) noted, fearing bodily sensations does
not mean that a person will necessarily be better at detecting interoceptive cues. On the
other hand, individuals with panic have heightened anxiety sensitivity (see Chapter 4)
and greater vigilance for the physical sensations associated with anxiety (e.g., Kroeze
& van den Hout, 2000a; Schmidt, Lerew, & Trakowski, 1997). We can conclude from
this that panic is characterized by a heightened vigilance and responsiveness to specific
physical symptoms linked to a core fear but it is unclear whether individuals with panic
disorder are better at detecting changes in their physical state.
Catastrophic Interpretations
A key feature of panic episodes is the tendency to interpret the occurrence of certain
bodily sensations in terms of an impending biological (e.g., death), mental (i.e., insanity),
or behavioral (e.g., loss of control) disaster (Beck, 1988; Beck & Greenberg, 1988; D.
M. Clark, 1986a). For example, individuals with panic disorder may interpret (a) chest
pain or a sudden increase in heart rate as sign of a possible heart attack, (b) shaking or
trembling as a loss of control, or (c) feelings of unreality or depersonalization as a sign
Panic Disorder
279
of mental instability or “going crazy.” Catastrophic misinterpretations are discussed
more fully in our review of the cognitive research.
Apprehension of Panic
Individuals with panic disorder report extreme distress, even terror, during panic attacks
and so quickly develop considerable apprehension about having future attacks. This fear
of panic is a distinguishing feature of the disorder and is included in DSM-IV-TR as a
diagnostic criterion (APA, 2000). Presence of fear and avoidance of panic attacks differentiates panic disorder from other anxiety disorders in which panic attacks occur but
the “fear of panic” is missing.
Extensive Safety Seeking and Avoidance
Safety-seeking behavior and avoidance of panic-related situations are common responses
to panic attacks and may be seen as coping strategies to prevent the impending disaster
(e.g., overwhelming panic, a heart attack, loss of control). Phobic avoidance is common
in panic disorder and is elicited by the anticipation of panic attacks in particular (Craske
& Barlow, 1988). The phobic situations associated with agoraphobia are quite variable
across individuals because the avoidance is elicited by the anticipation of panic attacks
and not by the situations themselves (White & Barlow, 2002). White et al. (2006)
reported that 98% of panic disorder cases have mild to severe situational avoidance,
90% experiential avoidance (i.e., use safety signals or thought strategies to withdraw
or minimize contact with a phobic stimulus), and 80% interoceptive avoidance (i.e.,
refusal of substances or activities that could produce the physical sensations associated
with panic). Furthermore, they found that severity of agoraphobic avoidance was predicted by elevated fear of physical symptoms of anxiety (i.e., anxiety sensitivity) and low
perceived control over threat. Together these findings indicate a close but complicated
relationship between panic attacks and the development of avoidance responses.
Perceived Lack of Control
Beck et al. (1985, 2005) noted that a striking characteristic of panic attacks is the feeling of being overwhelmed by uncontrollable anxiety. This apparent loss of control over
one’s emotions and the anticipated threat causes a fixation on the panicogenic sensations and a loss of capacity to use reason to realistically appraise one’s physical and
emotional state (Beck, 1988; see also Barlow, 2002).
Panic Distinct from Anxiety
McNally (1994) argues that panic should not be seen as an extreme form of anxiety
involving the anticipation of future threat but rather as an immediate “fight-or-flight”
response to perceived imminent danger. In the cognitive model of anxiety presented
in Chapter 2, panic attacks would fall within the “immediate fear response” (Phase
I), whereas apprehension about panic, avoidance, and safety-seeking would constitute
secondary processes (Phase II) that maintain a state of heightened anxiety about having
panic attacks.
280
TREATMENT OF SPECIFIC ANXIETY DISORDERS
Clinician Guideline 8.1
Panic attacks involve a sudden onset of intense fear of certain activated physical sensations
that are misinterpreted as indicating an imminent, even catastrophic, threat to one’s physical
or mental health. The misinterpretations of threat increases apprehension of and vigilance
regarding these physical symptoms, and leads to avoidance and safety-seeking responses to
reduce the possibility of future panic attacks.
Varieties of Panic
It is generally recognized that there are different types of panic attacks or episodes.
Table 8.2 presents five types of panic experience that may have distinct functional characteristics with implications for treatment.
Spontaneous and Situationally Cued Panic
DSM-IV-TR recognizes three types of panic. With spontaneous or unexpected (uncued)
panic attacks “the individual does not associate onset with an internal or external situational trigger (i.e., the attack is perceived as occurring spontaneous ‘out of the blue’),
[whereas] situationally bound (cued) panic attacks are defined as those that almost
invariably occur immediately on exposure to, or in anticipation of, the situational cue
or trigger” (APA, 2000, pp. 430–431). Examples of situationally cued panic include the
woman who always has a panic attack whenever she goes alone to a large department
store, the man who always has a panic episode whenever he drives outside the city limits, or the young person who panics at night when left alone in the house. Situationally
predisposed panic attacks are similar to situationally bound episodes but are not always
associated with the situational cues or do not necessarily occur immediately upon exposure to the situational trigger (APA, 2000). An example would be sometimes experi-
TABLE 8.2. Various Types of Panic Attacks
Type of panic attack
Description
Spontaneous panic
Unexpected (“out of the blue”) panic attacks that are not associated with
external or internal situational triggers (DSM-IV-TR; APA, 2000).
Situationally cued panic
Panic attacks that occur almost invariably with exposure or anticipated
exposure to a particular situation or cue (DSM-IV-TR; APA, 2000).
Nocturnal panic
A sudden awaking from sleep in which the individual experiences a state of
terror and intense physiological arousal without an obvious trigger (e.g., a
dream, nightmares).
Limited-symptom panic
A discrete period of intense fear or discomfort that occurs in the absence of a
real danger but involves less than four panic attack symptoms.
Nonclinical panic
Occasional panic attacks reported in the general population that often occur
in stressful or evaluative situations, involve fewer panic symptoms, and are
associated with less apprehension or worry about panic (McNally, 1994).
Panic Disorder
281
encing a panic attack while waiting in a bank line or attending the movies. As Taylor
(2000) noted, many factors can determine whether a situation increases the probability
of a panic attack including temperature, access to exits, crowding, familiarity, and the
like.
The distinction between uncued versus cued panic has important diagnostic implications in distinguishing panic disorder from other types of anxiety disorders. Although
panic attacks are present in the majority of anxiety disorders (over 80%), they are usually
associated with specific situations (e.g., anticipation of or exposure to a social encounter
in social phobia; see review by Barlow, 2002). For this reason DSM-IV-TR (APA, 2000)
requires the presence of at least two uncued or spontaneous panic attacks in order to
make a diagnosis of panic disorder. However, it can be difficult to determine if a panic
episode is entirely unexpected because we are dependent on the client’s retrospective
report and observational skills (McNally, 1994). The unexpectedness of panic probably
falls along a continuum, thereby making it difficult to assign panic attacks to a discrete
category of either expected or unexpected. Moreover, truly unexpected, uncued panic
attacks may be relatively infrequent, even in panic disorder (Brown & Deagle, 1992;
Street, Craske, & Barlow, 1989).
Clinician Guideline 8.2
Assessment for panic disorder should include a thorough evaluation of the frequency, severity, subjective probability, and contextual factors associated with spontaneous and situationally cued panic attacks.
Nocturnal Panic Attacks
Nocturnal panic attacks (NPs) are a frequent occurrence, with 25–70% of individuals
with panic disorder reporting at least one sleep panic attack, and 18–33% reporting frequent, recurrent NPs (Barlow, 2002; Craske & Rowe, 1997; Mellman & Uhde, 1989).
NPs, though phenomenologically similar to daytime panic attacks (Craske & Rowe,
1997), are characterized by an abrupt waking from sleep in a state of panic, especially
during the transition from Stage 2 to Stage 3 sleep (Barlow, 2002; Hauri, Friedman,
& Ravaris, 1989; Taylor et al., 1986). NPs are distinct from other sleep-related conditions such as night terrors, sleep apnea, sleep seizures, or sleep paralysis (Craske,
Lang, Aikins, & Mystkowski, 2005). There is some evidence that individuals with NPs
have more severe panic attacks than those with panic disorder without NPs, and many
patients with frequent NPs become fearful of sleep (Barlow, 2002; Craske & Rowe,
1997).
Craske and Rowe (1997; see also Aikins & Craske, 2007) proposed that the same
cognitive factors responsible for panic attacks in wakefulness are implicated in NPs.
Thus fear of change in physical state during sleep or relaxation, heightened vigilance
for and perception of changes in bodily state, and catastrophic appraisal of physiological changes immediately upon waking are considered important in the pathogenesis of
NPs. In NP distress about sleep and relaxation may reflect a fear of losing vigilance
for bodily changes during sleep (Aikins & Craske, 2007). In support of this cognitive-
282
TREATMENT OF SPECIFIC ANXIETY DISORDERS
behavioral explanation, studies have found an increase in physiological changes in the
minutes before panicky awaking (Hauri et al., 1989; Roy-Byrne, Mellman, & Uhde,
1988) and experimental manipulation of individuals’ expectations and interpretations
of the physiological arousal symptoms associated with sleep can influence their level
of anxiety and presence of panic attacks upon abrupt awaking (Craske et al., 2002;
see also Craske & Freed, 1995, for similar results). In addition Craske et al. (2005)
reported significant posttreatment gains at 9-month follow-up in a sample of panic
disorder patients with recurrent NPs who were offered 11 sessions of CBT. NPs, then,
are common in panic disorders and can be accommodated within the cognitive perspective.
Limited-Symptom Panic
DSM-IV-TR recognizes that limited-symptom attacks are common in panic disorder
and are identical to full-blown attacks except they involve fewer than 4 of 13 symptoms
(APA, 2000). The usual profile is for individuals to experience full-blown panic attacks
interspersed with frequent minor attacks, with both showing similar functional and
phenomenological characteristics (Barlow, 2002; McNally, 1994).
Nonclinical Panic
Contrary to expectations, panic attacks are actually quite common in the general
population. Questionnaire studies indicate that over one-third of nonclinical young
adults experience at least one panic attack within the past year (Norton, Dorward, &
Cox, 1986; Norton, Harrison, Hauch, & Rhodes, 1985), but only 1–3% report three
or more panic attacks in the last 3 weeks (i.e., Salge et al., 1988). Unexpected panic
attacks are less common, ranging from 7 to 28%, and far fewer (approximately 2%)
meet diagnostic criteria for panic disorder (Norton et al., 1986; Telch, Lucas, & Nelson, 1989). Structured interviews produce much lower rates (i.e., 13%) of nonclinical
panic (Brown & Deagle, 1992; Eaton, Kessler, Wittchen, & Magee, 1994; Hayward
et al., 1997; Norton, Cox, & Malan, 1992). However, the infrequent panic attacks of
infrequent nonclinical panickers are less severe, less pathological, and more situationally predisposed than the unexpected, “crippling” attacks found in diagnosable panic
disorder (Cox, Endler, Swinson, & Norton, 1992; Norton et al., 1992; Telch et al.,
1989), leading to the possibility that a history of infrequent panic attacks might be
a possible risk factor for panic disorder (e.g., Antony & Swinson, 2000a; Brown &
Deagle, 1992; Ehlers, 1995).
Clinician Guideline 8.3
The dimensional quality to panic attacks should be recognized when assessing this clinical
phenomenon. Clients should be evaluated for past and current experiences with less severe,
“partial” panic episodes as well as the occurrence of nocturnal panic attacks. An exclusive
focus on “full-blown” panic attacks may not capture the total impact of panic experiences
on individual clients.
Panic Disorder
283
Agoraphobic Avoidance
Agoraphobia is the avoidance or endurance with distress of “places or situations from
which escape might be difficult (or embarrassing) or in which help may not be available in the event of having a panic attack or panic-like symptoms” (DSM-IV-TR; APA,
2000, p. 432). The anxiety usually leads to pervasive avoidance of a variety of situations such as being at home alone, crowds, department stores, supermarkets, driving,
enclosed places (e.g., elevators), open spaces (e.g., crossing bridges, parking lots), theaters, restaurants, public transportation, air travel, and the like. In some cases agoraphobia is mild and confined to a few specific places, whereas for others it is more severe
in which a “safe zone” may be defined around the home with travel outside this zone
highly anxiety-provoking (Antony & Swinson, 2000a). In extreme cases, the person
may be completely housebound.
Panic attacks most often precede the onset of agoraphobia (Katerndahl & Realini,
1997; Thyer & Himle, 1985) and individuals with panic disorder are more likely to
develop agoraphobic avoidance to situations associated with the first panic attack (Faravelli, Pallanti, Biondi, Paterniti, & Scarpato, 1992). Furthermore, the development of
agoraphobic avoidance is less dependent on the frequency and severity of panic attacks
and more likely due to high anticipatory anxiety about the occurrence of panic, elevated
anxiety sensitivity, diminished sense of control over threat, and a tendency to use avoidance as a coping strategy (Craske & Barlow, 1988; Craske, Rapee, & Barlow, 1988;
Craske, Sanderson, & Barlow, 1987; White et al., 2006). The close association between
panic attacks and agoraphobia is also confirmed by the low prevalence of agoraphobia
without panic disorder (AWOPD). In the NCS-R AWOPD had a 12-month prevalence
rate of only 0.8% compared to 2.7% for panic disorder (Kessler et al., 2005) and rates
among treatment-seeking samples may be even lower because individuals with AWOPD
may be less likely to seek professional treatment (e.g., Eaton, Dryman, & Weissman,
1991; Wittchen, Reed, & Kessler, 1998). Although relatively rare, AWOPD may be
more severe and associated with less favorable treatment outcome than panic disorder,
but the studies are divided on whether it is characterized by greater impaired functioning (Buller, Maier, & Benkert, 1986; Buller et al., 1991; Ehlers, 1995; Goisman et al.,
1994; Wittchen et al., 1998).
Clinician Guideline 8.4
Expect some form of agoraphobic avoidance in most cases of panic disorder. It can vary from
mild, even fluctuating, forms of situational avoidance to severe cases of being housebound.
The clinician should adopt a broad, dimensional assessment perspective, with a focus on
recording the variety of situations, body sensations, feelings, and experiences that the client
avoids.
Diagnostic Features
Table 8.3 presents the DSM-IV-TR (APA, 2000) diagnostic criteria for panic disorder.
There are three possible diagnoses relevant to panic disorder; panic disorder without agoraphobia (300.01), panic disorder with agoraphobia (300.21), and agoraphobia
284
TREATMENT OF SPECIFIC ANXIETY DISORDERS
TABLE 8.3. DSM-IV Diagnostic Criteria for Panic Disorder
Criterion A. Both (1) and (2):
(1) recurrent unexpected Panic Attacks (i.e., at least two)
(2) at least one of the attacks has been followed by 1 month (or more) of one (or more) of the
following:
(a) persistent concern about having additional attacks
(b) worry about the implications of the attack or its consequences (e.g., losing control, having a
heart attack, “going crazy”)
(c) a significant change in behavior related to panic attacks
Criterion B.
Presence of agoraphobia is necessary for a diagnosis of Panic Disorder with Agoraphobia (300.21)
or absence of agoraphobia for a diagnosis of Panic Disorder without Agoraphobia (300.01)
Criterion C.
The Panic Attacks are not due to the direct physiological effects of a substance (e.g., drug of abuse,
a medication) or a general medical condition (e.g., hyperthyroidism)
Criterion D.
The Panic Attacks are not better accounted for by another mental disorder, such as Social Phobia
(e.g., occurring on exposure to feared social situations), Specific Phobia (e.g., on exposure to a
specific phobic situation), Obsessive–Compulsive Disorder (e.g., on exposure to dirt in someone
with an obsession about contamination), Posttraumatic Stress Disorder (e.g., in response to stimuli
associated with a severe stressor), or Separation Anxiety Disorder (e.g., in response to being away
from home or close relatives).
Note. From American Psychiatric Association (2000). Copyright 2000 by the American Psychiatric Association.
Reprinted by permission.
without a history of panic disorder (AWOPD; 300.22). The first two diagnoses are distinguished on the basis of presence or absence of situational avoidance. If a more inclusive definition of agoraphobic avoidance is used to include experiential and interoceptive
(internal) cues (White et al., 2006), then practically no one would receive a diagnosis of
panic disorder without agoraphobia.
Psychiatric Comorbidity
Panic disorder is associated with a high rate of diagnostic comorbidity. Based on a large
clinical sample (N = 1,127), Brown, Campbell, et al. (2001) found that 60% of individuals with a principal diagnosis of panic disorder with agoraphobia (n = 360) had at least
one other Axis I disorder. The most common comorbid conditions were major depression (23%), GAD (22%), social phobia (15%), and specific phobia (15%). PTSD (4%)
and OCD (7%) were relatively less common comorbid disorders. In the NCS 55.6%
of individuals with lifetime panic disorder met criteria for lifetime major depression,
whereas only 11.2% of those with lifetime major depression were comorbid for lifetime
panic disorder (Roy-Byrne et al., 2000). Panic disorder is more severe in those with
comorbid major depression (Breier, Charney, & Heninger, 1984). In terms of temporal
relationships, another anxiety disorder is more likely to precede panic with or without
agoraphobia (Brown, DiNardo, Lehman, & Campbell, 2001; Newman et al., 1996).
Panic Disorder
285
Substance abuse is also common in panic disorder (e.g., Sbrana et al., 2005). Results
of the National Epidemiologic Survey on Alcohol and Related Conditions (N = 43,093
respondents) indicate that panic disorder with agoraphobia and GAD were more likely
associated with a substance use disorder than other mood and anxiety disorders (Grant
et al., 2004). Rates of Axis II personality disorders range from 25 to 75%, with particular concentration in the Cluster C disorders (Diaferia et al., 1993; Renneberg, Chambless, & Gracely, 1992). Presence of borderline, dependent, schizoid, or schizotypal personality disorder by age 22 significantly predicted elevated risk for panic disorder by
age 33 (Johnson, Cohen, Kasen, & Brook, 2006). This finding is consistent with the
observed trend for nonpanic conditions to precede the development of panic disorder
when individuals have multiple diagnoses (Katerndahl & Realini, 1997).
Clinician Guideline 8.5
Presence of comorbid conditions, especially major depression, GAD, substance abuse, and
personality disorder, should be determined when conducting a diagnostic evaluation for
panic disorder.
Panic and Suicide Attempts
Although findings from the ECA suggested that individuals with panic disorder were
2.5 times more likely to attempt suicide than individuals with other psychiatric conditions (Weissman, Klerman, Markowitz, & Ouellete, 1989), later studies contradicted this finding, showing that suicide attempts are practically nonexistent in panic
disorder (e.g., Beck, Steer, Sanderson, & Skeie, 1991; Swoboda, Amering, Windhaber,
& Katschnig, 2003). More recently Vickers and McNally (2004) reanalyzed the
NCS data set and concluded that any suicide attempts in panic disorder were due to
psychiatric comorbidity and that panic itself did not directly increase risk for suicide
attempts.
Increased Medical Morbidity and Mortality
A number of medical conditions are elevated in panic disorder such as cardiac disease,
hypertension, asthma, ulcers, and migraines (Rogers et al., 1994; Stewart, Linet, &
Celentano, 1989). Panic sufferers are more likely to first seek medical evaluation of
their symptoms than attend a mental health setting (e.g., Katerndahl & Realini, 1995).
A significant number of individuals with cardiac complaints (9–43%) have panic disorder (Barsky et al., 1994; Katon et al., 1988; Morris, Baker, Devins, & Shapiro, 1997).
Moreover, higher rates of cardiovascular disease, even fatal ischemic heart attacks, have
been found in men with panic disorder (Coryell, Noyes, & House, 1986; Haines, Imeson, & Meade, 1987; Weissman, Markowitz, Ouellette, Greenwald, & Kahn, 1990). In
addition postmenopausal women who experience full-blown panic attacks have a threefold increased risk of coronary heart disease or stroke (Smoller et al., 2007). In a recent
cohort study based on analysis of the British General Practice Research Database, men
and women with panic disorder had a significantly higher incidence of coronary heart
286
TREATMENT OF SPECIFIC ANXIETY DISORDERS
disease and those younger than 50 years of age had a higher incidence of myocardial
infarction (Walters, Rait, Petersen, Williams, & Nazareth, 2008). Mitral valve prolapse
(MVP), a malformation of the leaflets of the heart’s mitral valve that causes symptoms
like chest pain, tachycardia, faintness, fatigue, and anxiety (see Taylor, 2000), is twice
as common in individuals with panic disorder as in nonpanic controls (Katerndahl,
1993). However, most individuals are asymptomatic and not at high risk for serious
health consequences (Bouknight & O’Rourke, 2000), so there is no clinical significance
in distinguishing panic patients with or without the condition (Barlow, 20002).
Panic disorder is associated with higher mortality rates possibly due to elevated risk
of cardiovascular and cerebrovascular diseases, especially in men with panic disorder
(Coryell et al., 1986; Weissman et al., 1990). Moreover, panic disorder and respiratory
diseases such as asthma (Carr, Lehrer, Rausch, & Hochron, 1994) and chronic obstructive pulmonary disease (Karajgi, Rifkin, Doddi, & Kolli, 1990) show a high rate of
incidence, although these diseases usually precede the onset of panic episodes. Panic
disorder is only diagnosed when there is clear evidence that the patient holds exaggerated negative beliefs about the dangerousness of unpleasant but harmless sensations like
breathlessness (Carr et al., 1994; Taylor, 2000).
There are a number of medical conditions that can produce physical symptoms
similar to panic disorder. These include certain endocrine disorders (e.g., hypoglycemia,
hyperthyroidism, hyperparathyroidism), cardiovascular disorders (e.g., mitral valve prolapse, cardiac arrhythmias, congestive heart failure, hypertension, myocardial infarction), respiratory disease, neurological disorders (e.g., epilepsy, vestibular disorders),
and substance use (e.g., drug/alcohol intoxication, or withdrawal) (see Barlow, 2002;
Taylor, 2000, for further discussion). Again, presence of these disorders does not automatically exclude the possibility of diagnosing panic disorder. If panic attacks precede
the disorder, occur outside the context of substance use, or the physical symptoms are
misinterpreted in a catastrophic fashion, than a diagnosis of comorbid panic disorder
should be considered in those with a medical condition (DSM-IV-TR; APA, 2000; Taylor, 2000). Other characteristics such as onset of panic attacks after age 45, presence of
unusual symptoms such as loss of bladder or bowel control, vertigo, loss of consciousness, slurred speech, and the like, and brief attacks that stop abruptly suggest a general
medical condition or substance use may be causing the panic (DSM-IV-TR; APA, 2000;
see Taylor, 2000). It is possible that physiological irregularities and ill health experiences could contribute to a heightened sensitivity to body sensations in panic disorder
(e.g., Hochn-Saric et al., 2004). For example, Craske, Poulton, Tsao, and Plotkin (2001)
found that experience with respiratory ill health or disturbance during childhood and
adolescence predicted the subsequent development of panic disorder with agoraphobia
at 18 or 21 years. Thus medical conditions can play either a contributing cause and/or
effect role in many cases of panic disorder.
Clinician Guideline 8.6
Most individuals with panic disorder have sought medical consultation prior to referral to
mental health services. However, a thorough medical examination should be obtained in
cases where a self-referral was made in order to rule out a co-occurring medical condition
that might mimic or exacerbate panic symptoms.
Panic Disorder
287
Descriptive Characteristics
Epidemiological studies indicate that panic disorder with or without agoraphobia have
1-year prevalence rates ranging from 1.1 to 2.7% and lifetime prevalence rates of 2.0–
4.7% (Eaton et al., 1991; Kessler et al., 1994; Kessler, Berglund, et al., 2005; Kessler,
Chiu, et al., 2005; Offord et al., 1996). This makes panic disorder second only to OCD
as the least common of the anxiety disorders discussed in this volume. As expected,
prevalence of panic disorder is much higher in primary care settings than in the general
population (Katon et al., 1986; Olfson et al., 2000). Moreover, there do not appear to be
significant ethnic differences in the prevalence of panic disorder (e.g., Horwath, Johnson, & Hornig, 1993; Kessler et al., 1994), although cultural factors do influence which
panic symptoms are more commonly reported and how they are labeled (see discussion
by Barlow, 2002; Taylor, 2000).
Panic attacks as well as panic disorder with or without agoraphobia are approximately twice as common in women as in men (Eaton et al., 1994; Gater et al., 1998;
Kessler et al., 1994). Moreover, agoraphobia may be particularly gendered, with women
representing approximately 75% of the agoraphobic population (Bourdon et al., 1988;
Yonkers et al., 1998). Panic disorder appears to take a more severe course in women
as indicated by more severe agoraphobic avoidance, more catastrophic thoughts, more
threatening interpretations of bodily sensations, and higher recurrence of panic symptoms (Turgeon, Marchand, & Dupuis, 1998; Yonkers et al., 1998). Women in general
may show a heightened tendency to report more physical symptoms, fear, and panic in
response to acute distress (Kelly, Forsyth, & Karekla, 2006). Furthermore, it is possible that increased panic disorder and agoraphobic avoidance in women is linked to a
higher rate of childhood physical and sexual abuse which could lead to increased hypervigilance and overpredictions of threat (Stein, Walker, et al., 1996). Craske (2003),
however, notes that the main difference between men and women is in their reliance on
avoidance rather than in the number of reported panic attacks, which could be due to
socialization into the traditional feminine gender role.
Panic disorder usually begins in young adulthood with the ECA reporting a mean
onset age of 24 years (Burke, Burke, Regier, & Rae, 1990) and 75% of panic disorder
cases having first onset by age 40 in the NCS-R survey (Kessler, Berglund, et al., 2005).
Despite a relatively early onset, there is usually considerable delay between onset and
first treatment contact. In the NCS-R a median duration of 10 years occurred between
onset of panic disorder and first treatment contact (Wang, Berglund, et al., 2005).
Despite lengthy delays in seeking treatment, the vast majority of individuals with panic
disorder eventually do make treatment contact (Wang, Berglund, et al., 2005).
Like other anxiety disorders, onset of panic is often associated with stressful life
events such as separation, loss or illness of significant other, being a victim of an assault,
financial problems, work difficulties, personal health problems, unemployment, and the
like (e.g., Faravelli & Pallanti, 1989; Franklin & Andrews, 1989; Pollard, Pollard, &
Corn, 1989). In other studies a high incidence of past childhood sexual and physical
abuse has been found in panic disorder and agoraphobia, especially among women
(Pribor & Dinwiddie, 1992; Saunders, Villeponteaux, Lipovsky, Kilpatrick, & Veronen,
1992; Stein et al., 1996), although it is still uncertain whether rates of adverse early
childhood events are any higher in panic disorder compared to major depression or even
schizophrenia (Friedman et al., 2002). In an analysis of the NCS data set women with
panic disorder without comorbid PTSD were six times more at risk of having child-
288
TREATMENT OF SPECIFIC ANXIETY DISORDERS
hood physical or sexual abuse, and individuals with comorbid panic and PTSD were
significantly more likely to have survived rape (43%) than those with panic alone (7.5%)
(Leskin & Sheikh, 2002). These findings indicate that lifetime trauma may act as a risk
factor for panic disorder, especially in women. Moreover, social environmental factors
may also affect clinical course, with factors such as childhood separation, lower socioeconomic status, and marital breakup significant predictors of poor outcome 7 years
after initial treatment (Noyes et al., 1993).
Relationship problems may be more common in panic disorder than in other conditions, both as a contributing cause and a consequence of the disorder (Marcaurelle,
Bélanger, & Marchand, 2003). However, the empirical evidence is inconsistent in
whether panic disorder with agoraphobia is associated with more marital problems and
quality of marital relationship at pretreatment is not a significant predictor of treatment
prognosis (Marcaurelle et al. 2003).
If left untreated, panic disorder typically takes a chronic course with only 12% of
patients achieving complete remission after 5 years (Faravelli, Paterniti, & Scarpato,
1995). In a 1-year prospective study Ehlers (1995) found that 92% of panic patients continued to experience panic attacks and 41% of the initially remitted patients relapsed.
However, in an 11-year follow-up of 24 patients with panic disorder treated in an 8-week
clinical trial of imipramine, alprazolam, or placebo, 68% had no panic attacks over the
follow-up period and 90% showed no or only mild disabilities (Swoboda et al., 2003).
This suggests that with treatment, the long-term prognosis for panic disorder may be
more optimistic.
Panic disorder is also associated with significant functional impairment and decrements in quality of life, especially when comorbid with depression (Massion, Warshaw,
& Keller, 1993; Roy-Byrne et al., 2000; Sherbourne et al., 1996). Furthermore, greater
functional impairment can significantly increase the likelihood of panic recurrence in
previously recovered individuals (Rodriguez, Bruce, Pagano, & Keller, 2005). In a metaanalytic review of 23 quality of life studies, panic disorder was similar to the other anxiety disorders in showing significant decrements in physical health, mental health, work,
social functioning, and family functioning (Olatunji et al., 2007), although poor subjective quality of life is worse in major depression than in panic disorder (Hansson, 2002).
Panic disorder with agoraphobia can be a costly disorder both in terms of human
suffering and increased burden on health care services (e.g., Eaton et al., 1991). In the
NCS-R panic disorder and PTSD had the highest annual rates of mental health utilization compared to the other anxiety disorders, and panic disorder had a much higher rate
of accessing general medical care (Wang, Lane, et al., 2005; see also Deacon, Lickel,
& Abramowitz, 2008). The health care costs associated with panic disorder, then, are
substantial. The number of annual medical visits by individuals with panic disorder is
seven times that of the general population, resulting in an annual medical cost that is
twice the American population average (Siegel, Jones, & Wilson, 1990).
Clinician Guideline 8.7
Negative life events, past and current stressors, negative coping style, and psychosocial
impairment will have a significant impact on the course of panic disorder. The clinician
must take these factors into consideration during assessment and treatment of panic.
Panic Disorder
289
COGNITIVE THEORY OF PANIC DISORDER
Description of the Model
Panic attacks are an immediate fear response and so the psychogenic processes primarily responsible for the onset and persistence of panic occur within Phase I of the cognitive model depicted in Chapter 2 (see Figure 2.1). The key cognitive processes of panic,
then, occur at an automatic level of activation. The Phase II processes, representing
deliberate, elaborative efforts to cope with heightened anxiety and the anticipation of
panic, are secondary contributors to the persistence of the disorder. In this way the cognitive basis of panic is entirely different from that of GAD where Phase II processes play
a more critical role in the disorder. The cognitive model of panic was first articulated in
the mid- to late 1980s by Beck and colleagues (Beck, 1988; Beck et al., 1985; Beck &
Greenberg, 1988; D. M. Clark & Beck, 1988) and further elaborated by D. M. Clark
(1986a, 1988, 1996; D. M. Clark et al., 1988). Derived from these earlier accounts,
Figure 8.1 illustrates the current cognitive explanation for panic based on the generic
cognitive model (i.e., Figure 2.1). It should be noted that the cognitive model was formulated to explain the pathogenesis of recurrent panic attacks or panic disorder. It has less
relevance for understanding the occasional panic attacks found in the general population or the occurrence of nonfearful panic-like somatic symptoms prominent in medical
settings (D. M. Clark, 1997; see also Eifert, Zvolensky, & Lejuez, 2000).
Activation and Attention
Changes in internal states such as the occurrence or intensification of certain physical
sensations (e.g., chest tightness, breathlessness, increased heart rate, nauseous feelings)
or mental processes (e.g., mind goes blank, sense of derealization) are the primary
triggers for panic attacks. In most cases of recurrent panic external stimuli or situations will become triggers, but only because they have the capacity to elicit bodily
sensations that are perceived as threatening (D. M. Clark, 1986a). In our case example
Helen’s primary trigger for panic was a sense of breathlessness. However, she perceived
changes in her breathing whenever she was in novel situations and so unfamiliar settings such as travel outside her community, visiting new people, and so on began to
trigger heightened anxiety and anticipated panic. Naturally she started avoiding these
situations because they elicited the threatening somatic sensation, breathlessness. In
severe agoraphobia a wide variety of external situations can trigger panic but only
because they elicit some feared internal state. The physical and mental processes that
are misperceived as an imminent threat are most often due to anxiety, and less often
due to other emotional states, stress, physical exertion, ingestion of substances with
caffeine or other chemicals, or even the natural ebb and flow of physiological function
(D. M. Clark, 1986a, 1988, 1996).
As noted in Figure 8.1, individuals prone to panic are oriented toward selectively
attending to internal somatic or mental processes. They are hypervigilant for the experience of these sensations and focus their attention on any change in functioning that
might seem abnormal (Beck, 1988). The orienting mode in panic disorder is primed
toward rapid detection of interoceptive cues that could represent an immediate and
imminent danger to survival. This early detection process is automatic and nonconscious, resulting in a hypersensitivity to bodily sensations.
290
TREATMENT OF SPECIFIC ANXIETY DISORDERS
ACTIVATING CUES, TRIGGERS
(Internal bodily sensation, physiological change)
ORIENTING MODE
(Selective attention to
interoceptive cues)
PHASE I
PROCESSES
PRIMAL MODE
Activation of
Physiological or Mental
Threat Schemas
BIASED COGNITIVE
PROCESSING
AUTONOMIC AROUSAL
(Physical symptoms intensify)
AUTOMATIC THOUGHTS
IMMEDIATE
(Exaggerated, catastrophic
threat appraisals)
DEFENSIVE RESPONSE
(Escape)
PHASE II
PROCESSES
LOSS OF
PERCEIVED LOSS
REAPPRAISAL
WORRY, APPREHENSION
OF CONTROL
CAPABILITY
ABOUT ANXIETY, AND
FUTURE PANIC
SEARCH FOR
AVOIDANT
SAFETY
COPING
FIGURE 8.1. Cognitive model of panic disorder.
Panic Disorder
291
Schematic Activation
In the cognitive model recurrent panic attacks occur because of the activation of enduring schemas (beliefs) about the dangerousness of particular physiological changes.
These physiological and mental threat schemas are consistent with the primal threat
mode that dominates in anxiety. Some examples of panicogenic schemas are “My heart
palpitations and chest pain might be signs that I am having a potentially fatal heart
attack,” “My episodes of breathlessness could lead to suffocation and death,” “Dizzy
spells might be caused by a brain tumor,” “This feeling of nausea could cause me to be
violently sick and vomit controllably,” “Feeling tense and a little nervous could lead to
loss of control and overwhelming panic,” or “I could lose control of my emotions and
go crazy.” The physiological and mental threat schemas predispose certain individuals to experience recurrent panic attacks and involve themes of personal vulnerability,
helplessness, the dangerousness of particular internal states, and the rapid escalation of
anxiety (Beck & Greenberg, 1988).
There is a close association between specific bodily or mental sensations and the
panicogenic beliefs that underlie the catastrophic misinterpretation of interoceptive
cues. Table 8.4 presents connections between some common internal sensations and the
corresponding physiological or mental threat schemas often seen in panic disorder.
Panic-relevant schemas are quite idiosyncratic and can be highly variable so that
some individuals will hold exaggerated threat beliefs for only certain internal sensations,
others will misinterpret a wide range of somatic and mental phenomena as threatening,
TABLE 8.4. Links between Panic-Relevant Internal Sensations and Their Corresponding
Physiological or Mental Threat Schema
Internal sensation
Physiological/mental threat schema
Chest tightness, pain, heart palpitations
Belief of vulnerability to heart attacks.
Breathlessness, smothering sensation,
irregular breathing
Belief of possible suffocation and death.
Dizziness, lightheadedness, faintness
Belief of losing control and doing something embarrassing or
of going crazy, or presence of a brain tumor.
Nausea, abdominal cramps
Belief of vomiting uncontrollably.
Numbness, tingling in extremities
Belief in the possibility of having a stroke or losing one’s
sanity.
Restless, tense, agitation
Belief that these symptoms are an indication of losing control
which could lead to a severe panic attack and eventual loss of
function.
Feel shaky, trembling
Belief these symptoms indicate a loss of control and are often a
precursor to severe panic.
Forgetful, inattentive, loss of
concentration
Belief of losing control over one’s mental functioning and
ending up losing cognitive function.
Feelings of unreality, depersonalization
Belief that these symptoms may indicate a seizure or of going
crazy.
Note. Based on Taylor (2000).
292
TREATMENT OF SPECIFIC ANXIETY DISORDERS
and still others may show a shift in which only certain physiological threat schemas
are dominant at any particular time (D. M. Clark, 1986a). Acquisition of particular
physiological threat schemas (e.g., “Heart palpitations are dangerous”) will depend on
prior learning history and the circumstances associated with the first panic attack (D.
M. Clark, 1997). For example, it is common for individuals who experience chest pain
to first go to emergency departments where they receive a full cardiac assessment. Such
experiences can reinforce beliefs that “chest pain represents a highly imminent danger
of heart attack and possible death.” It is obvious how such experiences can lead to
hypervalent schemas about the dangerousness of chest pain and the pathogenesis of
panic disorder.
In order to activate physiological and mental threat schemas, the corresponding
internal sensations must reach a certain threshold of intensity (Beck, 1988). For example, Helen did not experience heightened anxiety until her sense of breathlessness was
sufficiently intense that she began to wonder if she was getting enough air. Furthermore,
once schematic activation occurs, the main consequence is the catastrophic misinterpretation of the internal sensation. Once Helen’s beliefs about the danger of breathlessness,
suffocation, and lack of oxygen were activated by attention to her respiratory sensations,
she made a rapid, automatic catastrophic misinterpretation. “There is something wrong
with my breathing and I am not getting enough oxygen; I could suffocate to death.”
Thus the cognitive basis of the catastrophic misinterpretation of bodily sensations is
the activation of prepotent and enduring threat-oriented schemas about the imminent
danger associated with certain somatic or mental sensations.
Catastrophic Misinterpretation of Internal Sensations
The central cognitive process in the persistence of panic is the catastrophic misinterpretation of somatic or mental sensations (see Beck, 1988; Beck et al., 1985; D. M.
Clark, 1986a). Often the catastrophic outcome associated with physical sensations is
death caused by heart attack, suffocation, seizure, or the like. However, the imagined
catastrophe can also involve a loss of control that leads to insanity (i.e., “I’ll go crazy”),
or acting in an embarrassing or humiliating manner in front of others. In addition,
fear of panic attacks can be so intense that the catastrophe might be the possibility of
experiencing another severe full-blown panic episode. Whatever the actual nature of
the dreaded catastrophe, the sensations are misinterpreted as representing an imminent
physical or mental disaster (D. M. Clark, 1988). In order to precipitate panic, the catastrophic threat must be perceived as imminent; if the misinterpretation is merely exaggerated threat, then anxiety rather than panic will be roused (Rachman, 2004). The
occurrence of the catastrophic misinterpretation is the cognitive basis for the remaining
processes that contribute to panic disorder (see Figure 8.1). In cued or situational panic
attacks sensations associated with heightened anxiety are misinterpreted, whereas in
spontaneous (uncued) panic attacks the sensations arise from a variety of nonanxious
sources (e.g., exercise, stress, emotional reactions). D. M. Clark (1988) argued that the
catastrophic misinterpretation of bodily sensations is necessary for the production of
a panic attack and represents an enduring cognitive trait (vulnerability) that is evident
even when individuals with panic disorder are not anxious.
As a Phase I process that is elicited by activation of panicogenic schemas, the catastrophic misinterpretation is an involuntary, automatic, and rapid response to the detec-
Panic Disorder
293
tion of certain internal sensations. D. M. Clark (1988) argued that catastrophic misinterpretations can be rapid and outside conscious awareness so the panic attack appears
spontaneous. Once activated the panicogenic schemas and catastrophic misinterpretations tend to dominant the information-processing apparatus during panic. It is highly
biased, giving processing priority to danger cues and minimizing or ignoring schemaincongruent safety information.
Symptom Intensification and Defense
The catastrophic misinterpretation of bodily sensations will cause an intensification of
the feared internal sensations by heightening vigilance and an internal focus on interoceptive cues (Beck, 1988; D. M. Clark, 1997). A vicious cycle occurs in which the escalating intensity of the physiological or mental sensation further reinforces the misinterpretation that indeed a physical or mental disaster is imminent. Helen, for example, would
notice that her breathing was a little irregular and felt like she was not getting enough air.
Her initial appraisal “I am not breathing normally, I don’t think I am getting enough air”
(exaggerated threat interpretation) led to an increased focus on her breathing. She would
breathe deeply and try to establish a more controlled breathing rate. But the heightened
focus on her breathing intensified her sense of breathlessness (symptom intensification)
that in turn deepened her conviction that her respiratory problem was getting worse and
even more likely to lead to suffocation (catastrophic misinterpretation).
In addition to an automatic intensification of symptoms, the catastrophic misinterpretation will result in immediate attempts to escape. Again, efforts to escape are
viewed as an automatic and involuntary response to panicogenic schema activation and
the catastrophic misinterpretation of the bodily sensation. A person fearful of chest
tightness may quickly cease a particular activity when the somatic sensation is detected.
A patient who had a profound sense of derealization in his hotel room after driving in
heavy traffic in New York City immediately laid down and then had several alcoholic
drinks in order to relax. His responses were an attempt to escape from the sense of
derealization which he interpreted as a symptom of going crazy. When in a state of
heightened anxiety and panic, the escape response triggered by the catastrophic misinterpretation can occur quite automatically without deliberate, effortful planning. As
noted in Figure 8.1, the intensification of physical sensations and escape responses will
contribute to the continued activation of the panic-relevant schemas.
Loss of Reappraisal Capability
According to Beck (1988), the dissociation of the higher level reflective processes (Phase
II) from the automatic cognitive processing (Phase I) is a necessary condition for a
panic attack. Thus we consider the loss of reappraisal capability the central process at
the secondary, elaborative phase that is responsible for the persistence of anxiety and
panic. Activation of the physiological threat schemas and subsequent catastrophic misinterpretation of bodily sensations dominates information processing and inhibits the
panic-stricken patient’s ability to generate alternative, more realistic, and benign interpretations of the fearful sensations. If reappraisal of the perceived threat is possible, the
catastrophic misinterpretation would be challenged and the escalation into panic would
be thwarted.
294
TREATMENT OF SPECIFIC ANXIETY DISORDERS
This loss of reappraisal capability is clearly illustrated in a young man with panic
disorder who was fearful of sudden increases in his heart rate. On some occasions, such
as sitting at his computer, he would perceive an increase in heart rate that elicited the
apprehensive thought “Why is my heart racing?” His underlying physiological threat
schemas were “I am vulnerable to heart attacks,” “If I let my heart rate get too high,
I could have a heart attack,” and “After all, I do have a cardiac condition” (he had a
diagnosed congenital cardiac condition that was benign). Once activated, he generated a
catastrophic misinterpretation (“My heart is racing, I might be having a heart attack”).
At this point he was unable to generate an alternative explanation for this increased
heart rate, and so he became panicky. On other occasions, such as when working out
at the gym (as recommended by his physician), he would notice his heart rate increase,
wonder if it could be a sign of a cardiac problem, but immediately reappraised the sensations as due to the demands of his physical activity. One of the main objectives of cognitive therapy for panic is to improve the patient’s ability to reappraise fearful internal
sensations with more realistic, plausible, and benign alternative interpretations.
Other Secondary Elaborative Processes
As illustrated in Figure 8.1 there are a number of other secondary cognitive and behavioral processes that occur as a result of the dissociation of elaborative reasoning from
the automatic catastrophic threat appraisals. Beck et al. (1985) noted that a striking
characteristic of panic attacks is the experience of anxiety as an overwhelming and
uncontrollable state. The individual with recurrent panic attacks thinks of anxiety as a
rapidly escalating and uncontrollable experience that she learns to dread.
A second cognitive process at the elaborative phase is apprehension and worry about
mounting anxiety and the recurrence of panic attacks. The worry in panic disorder is
focused almost exclusively on panic attacks and the intolerance of heightened states of
anxiety. After a number of CT sessions, Helen’s panic attacks remitted. However, her
apprehension and worry over a possible relapse remained high. For example, she was
considering a change of jobs and a move to a new city but was very reluctant to make
any changes for fear it would heighten her anxiety and trigger a new round of panic
attacks.
With elaborative information processing dominated by perceptions of uncontrollable and escalating anxiety, constant apprehension and worry about panic, and loss of
higher order reflective reasoning to counter the domination of catastrophic thinking,
it is little wonder that the person with panic disorder deliberately turns to avoidance
and other safety-seeking strategies to exert better control over his negative emotional
state. However, there is now considerable evidence that agoraphobic avoidance actually contributes to the persistence and increased severity of panic disorder (see previous
discussion). Moreover, reliance on safety-seeking behaviors such as carrying anxiolytic
medication in case of emergency, being accompanied by a family member or friend, or
suppressing strong emotions and unwanted thoughts, can actually contribute to the persistence of panic by maintaining the person’s belief that certain internal sensations are
dangerous (D. M. Clark, 1997, 1999).
As can be seen from Figure 8.1, there is a strong reciprocal relationship between the
panic-relevant cognitive processes that occur early at the automatic, catastrophic inter-
Panic Disorder
295
pretation level and those that occur later at the secondary, elaborative phase. However,
the inability of secondary elaborative thinking to correct the automatic catastrophic
threat appraisals of bodily sensations accounts for the persistence of panic and the
development of panic disorder. Cognitive therapy of panic, then, focuses on redressing
the dissociation between the two levels of processing so that a more benign interpretation of previously feared interoceptive sensations is accepted.
EMPIRICAL STATUS OF THE COGNITIVE MODEL
The proposition that panic attacks are caused by the catastrophic misinterpretation of
bodily sensations has generally received strong empirical support from a large number of correlational and experimental studies conducted over the past two decades,
although inconsistencies and limitations have also been noted (for reviews, see Austin &
Richards, 2001; Casey, Oei, & Newcombe, 2004; D. M. Clark, 1996; Khawaja & Oei,
1998; McNally, 1994). In this section we expand our review of the empirical status of
the cognitive model to include additional cognitive processes that are important in the
pathogenesis of panic. Table 8.5 presents six hypotheses that capture the main tenets of
the cognitive model of panic (see Figure 8.1).
TABLE 8.5. Core Hypotheses of the Cognitive Model of Panic
Hypotheses
Statement
1. Interoceptive
hypersensitivity
Individuals with panic disorder will exhibit selective
attention to and greater vigilance for internal somatic and
mental sensations than individuals without panic disorder.
2. Schematic vulnerability
Panic-prone individuals will endorse more beliefs about the
dangerousness of specific physiological or mental sensations
than nonpanic comparison groups.
3. Catastrophic
misinterpretations
Panic attacks are characterized by a misinterpretation of
bodily or mental sensations as signifying an imminent
physical, mental, or social catastrophe. Production of
the catastrophic misinterpretation will increase panic
symptoms in panic disorder individuals, whereas correction
of the misinterpretation will prevent panic attacks.
4. Interoceptive amplification
The production of a catastrophic misinterpretation of
internal cues will heighten the intensity of the feared
sensations in panic but not in nonpanic states.
5. Dissociation
Individuals with panic disorder will exhibit diminished
ability to employ higher order reflective thinking to
generate more realistic and benign interpretations of
their fearful internal sensations compared to nonpanic
individuals.
6. Safety seeking
Avoidance and maladaptive safety-seeking behavior will
intensify anxiety and panic symptoms in those with panic
disorder relative to nonpanic controls.
296
TREATMENT OF SPECIFIC ANXIETY DISORDERS
Hypothesis 1. Interoceptive Hypersensitity
Individuals with panic disorder will exhibit selective attention to and greater vigilance for
internal somatic and mental sensations than individuals without panic disorder.
If panic disorder is characterized by heightened vigilance and response to bodily
sensations, at the very least one would expect individuals with panic disorder to report
greater response to physical sensations on questionnaire and interview measures. In
several studies individuals with panic disorder and agoraphobia scored significantly
higher on the Body Sensations Questionnaire (BSQ), which assesses fear of 17 physical and mental sensations common in anxiety and panic, compared to individuals
with other anxiety disorders or nonclinical control groups (e.g., Chambless & Gracely,
1989; Kroeze & van den Hout, 2000a; Schmidt et al., 1997). Similarly McNally et al.
(1995) found that individuals with panic reported more severe physical sensations than
nonclinical controls, with fear of dying, fear of heart attack, fear of losing control,
and tingling being the best discriminators. However, individuals with panic disorder
may have heightened discomfort intolerance, as indicated by a reduced ability to withstand unpleasant physical sensations and pain more generally (Schmidt & Cook, 1999;
Schmidt, Richey, & Fitzpatrick, 2006). Overall there is fairly consistent evidence that
individuals who experience recurrent panic attacks report greater sensitivity to physical
sensations and are more likely to interpret these symptoms negatively (see also Taylor,
Koch, & McNally, 1992).
Stronger support for the interoceptive hypersensitivity hypothesis comes from
experimental studies that induce physical sensations through various biological challenges such as hyperventilation, inhalation of CO2-enriched or O2-enriched air, lactate
infusion, and the like. A consistent finding across these experimental studies is that
panic disorder patients evidence a significantly greater subjective response to the sensations produced by the inductions as indicated by higher ratings on the intensity, severity, and anxiousness associated with the bodily sensations produced by the induction
manipulations (e.g., Antony, Coons, McCabe, Ashbaugh, & Swinson, 2006; J. G. Beck,
Ohtake, & Shipherd, 1999; Holt & Andrews, 1989; Rapee, 1986; Schmidt, Forsyth,
Santiago, & Trakowski, 2002; Zvolensky et al., 2004).
If panic disorder is characterized by increased vigilance for physical sensations, we
might expect panic disorder patients to demonstrate greater acuity or perception of their
physiological responding. A number of studies have investigated heart rate perception
in panic disorder. In an early study by Pauli et al. (1991) panic disorder individuals who
wore an ECG recorder over 24 hours did not report significantly more cardiac perceptions than healthy controls but significantly more self-reported anxiety was associated
with the perceptions. Moreover, heart rate acceleration occurred after cardiac perceptions that were associated with intense anxiety whereas cardiac perceptions associated
with no anxiety led to heart rate deceleration.
Some studies have used a “mental tracking” procedure in which individuals silently
count felt heartbeats without taking their pulse. Early findings suggested that individuals with panic disorder had better heartbeat perception than other patient groups or
nonclinical controls (e.g., Ehlers & Breuer, 1992; Ehlers, Breuer, Dohn, & Fiegenbaum,
1995), but a later reanalysis of pooled data across different studies found that accurate
heartbeat perception was more often evident in panic disorder compared to depressed
Panic Disorder
297
and normal controls but not when compared to patients with other anxiety disorders
(van der Does, Antony, Ehlers, & Barsky, 2000). Moreover, only a minority of the panic
disorder patients was classified as accurate perceivers (17%). Thus accurate heartbeat
perception appears to be a characteristic of having frequent episodes of clinical anxiety
as opposed to panic attacks per se.
An automatic, preconscious attentional processing bias for physical cue words
should be apparent if panic is characterized by hypervigilance for bodily sensations.
Lundh and colleagues (1999) found that panic disorder patients had significantly higher
Stroop interference effects to panic-related words than nonclinical controls at both a
subliminal and a supraliminal level but this biasing effect was also evident for interpersonal threat words. In addition the panic disorder group identified more panic-related
words presented at perceptual threshold (see also Pauli et al., 1997). Using a novel variant of the dot probe detection task in which response latency was assessed to a letter
preceded by a snapshot sample of ECG heart rate data or a moving line, Kroeze and van
den Hout (2000a) found evidence that the panic group was more fully attentive to the
ECG trials than the control group (see Kroeze & van den Hout, 2000b, for contrary
finding).
In a study involving 20 individuals with claustrophobia, those told to concentrate
on their bodily sensations while in an enclosed chamber reported significantly higher
fear and panic scores, and experienced a higher rate of panic attacks than individuals in
the control (distraction) group (Rachman, Levitt, & Lopatke, 1988). Strenuous physical exercise is a naturalistic situation that normally increases attention to physical state.
Furthermore, vigorous exercise increases blood lactate levels, which individuals with
panic might find less tolerable given their heightened reactivity to sodium lactate infusion (Liebowitz et al., 1985). So, one might expect panic patients to be less tolerant of
strenuous physical exercise. Interestingly, it appears that individuals with panic disorder
are able to engage in vigorous physical exercise without experiencing thoughts or feelings indicative of panic even though the exercise produces blood lactate levels that are
equal to or greater than those attained in lactate infusion studies (Martinsen, Raglin,
Hoffart, & Friis, 1998).
Although individuals with panic disorder may have greater physiological reactivity such as a elevated respiratory rate, heart rate, and blood pressure, and lower skin
temperature during biological provocations that induce bodily sensations (J. G. Beck et
al., 1999; Craske, Lang, Tsao, Mystkowski, & Rowe, 2001; Holt & Andrews, 1989;
Rapee, 1986; Schmidt et al., 2002), the physiological differences are relatively modest
and inconsistent across studies, with some even reporting negative results (Zvolensky
et al., 2004). On the other hand, differences in perceived intensity and distress of the
physical sensations produced by these biological challenges have been robust and quite
consistent across studies (e.g., J. G. Beck et al., 1999; Holt & Andrews, 1989; Rapee,
1986). In a recent study Story and Craske (2008) found that individuals at risk for panic
(high anxiety sensitivity and a history of panic attacks) reported significantly more
panic symptoms following false elevated heart-rate feedback than low-risk individuals,
even though there were no group differences in actual heart rate. Together these findings
provide strong evidence for the cognitive perspective on panic disorder, suggesting that
the main difference is in the perception and interpretation of physical changes rather
than in actual physiological responses.
298
TREATMENT OF SPECIFIC ANXIETY DISORDERS
In summary, there has been fairly consistent empirical support from self-report and
biological challenge experiments that panic disorder is characterized by a heightened
sensitivity or perceptual bias to physical sensations, even though they may not have
enhanced physiological reactivity (Ehlers, 1995). The findings of greater perceptual acuity for interoceptive cues (e.g., enhanced cardiac awareness), however, remains uncertain. Moreover, it is clear that contextual factors affect response to physical sensations
and their interpretation. When bodily sensations occur in unexpected or anxious situations, individuals with panic disorder are likely to be more vigilant and responsive to
changes in their physical or mental state.
Hypothesis 2. Schematic Vulnerability
Panic-prone individuals will endorse more beliefs about the dangerousness of specific
physiological or mental sensations than nonpanic comparison groups.
In their critical review of the cognitive perspective on panic disorder, Roth, Wilhelm, and Pettit (2005) noted that if individuals with panic disorder did not exhibit
enduring “catastrophic beliefs” when panic attacks are absent, then this would be problematic for the theory. According to the schema vulnerability hypothesis, individuals
with panic disorder are expected to exhibit stronger endorsement of thoughts, assumptions, and beliefs that reflect activation of physiological threat schemas than nonpanic
disorder individuals even in the absence of a panic attack. Unfortunately, very little
research has specifically focused on beliefs in panic disorder. Khawaja and Oei (1992)
developed the 50-item Catastrophic Cognitions Questionnaire to assess misinterpretations of the dangerousness of specific physical, emotional, and mental states but the
measure failed to differentiate panic from other anxiety disorders (Khawaja, Oei, &
Baglioni, 1994). Greenberg (1989) constructed the 42-item Panic Belief Questionnaire
(PBQ) to assess level of agreement to maladaptive panic-related beliefs. The PBQ had a
moderate correlation with the ASI (r = .55) and panic disorder patients scored higher on
the total score than a social phobia group, although the difference was not statistically
significant (Ball, Otto, Pollack, Uccello, & Rosenbaum, 1995). More recently Wenzel et
al. (2006) reported that the PBQ Physical Catastrophes subscale had strong correlations
with other panic symptom questionnaires and that scores on the measure declined significantly with treatment. Inspection of the PBQ item content indicates that only seven
items (17%) tap into beliefs about physical sensations. Thus at present we do not have
a self-report measure that specifically assesses the enduring physiological and mental
threat schemas proposed by the cognitive model.
Currently the strongest self-report evidence for the schema vulnerability hypothesis
comes from research on anxiety sensitivity (see discussion in Chapter 4). Even though
the ASI is not a belief measure per se, it does assess an enduring tendency to interpret
physical sensations in a threatening manner, which is relevant to the nature of preexisting physiological threat schemas. Evidence that individuals with panic disorder score
significantly higher than other anxiety groups, especially on the ASI Physical Concerns
subscale, and that high ASI scores predict response to biological challenge experiments
as well as development of panic attacks is entirely consistent with the schema vulnerability hypothesis for panic disorder. However, the same type of research that has been
conducted on the ASI needs to be extended to a specific panic belief measure like the
Panic Disorder
299
PBQ in order to determine if physiological and mental threat schemas play a critical role
in the development of panic disorder.
If beliefs about the threatening nature of internal states are preexisting cognitive
structures, then individuals prone to panic disorder should evidence biased processing of
panic-relevant information even during nonanxious or nonpanic states. In fact there is a
large body of information-processing research that is consistent with activation of physiological or mental threat schemas in panic disorder. Experiments employing the emotional Stroop task have shown that compared to nonclinical control groups, individuals
with panic disorder exhibit a specific color-naming interference for physical threat or
catastrophe words (Hayward et al., 1994; McNally et al., 1994; Teachman, Smith-Janik,
& Saporito, 2007) even at subliminal presentation rates (Lim & Kim, 2005; Lundh et
al., 1999). However, some studies have found that the interference effect in panic is evident for all threat words in general (Ehlers, Margraf, Davies, & Roth, 1988; Lundh et
al., 1999; McNally, Kaspi, Riemann, & Zeitlin, 1990) or even all emotionally valenced
words (Lim & Kim, 2005; McNally et al., 1992). A few studies have reported no specific
color-naming interference for physical threat words in panic disorder (Kampman, Keijsers, Verbraak, Näring, & Hoogduin, 2002; McNally et al., 1992). Nevertheless, the
general findings from the emotional Stroop experiments are consistent with the presence
of prepotent physiological and mental threat schemas in panic disorder.
Evidence has also been found for an interpretation bias for internal stimuli in panic
disorder. Harvey et al. (1993) found that panic disorder patients chose threat explanations for ambiguous interoceptive scenarios more than social phobics, although there
was no statistical significance between the groups in how often they made threat interpretations. In a covariation bias experiment Wiedemann, Pauli, and Dengler (2001)
found that individuals with panic disorder but not healthy controls overestimated the
association between emergency room pictures (i.e., panic-relevant stimuli) and a negative
consequence (i.e., harmless shock to forearm). However, this finding was not replicated
in a later study (Amrhein, Pauli, Dengler, & Wiedemann, 2005), although electrophysiological evidence for a covariation bias in panic disorder was found. Coles and Heimberg (2002) in their review concluded that panic disorder is characterized by an explicit
but not an implicit memory bias for threatening information, especially when deep processing is encouraged at the encoding stage. Moreover, the explicit memory bias may
be especially pronounced with physical threat information (Becker et al., 1994; Cloitre
et al., 1994; Pauli, Dengler, & Wiedermann, 2005), although others have failed to find
a specific memory bias (Baños et al., 2001; Lim & Kim, 2005). Finally, Teachman et
al. (2007) found that individuals with panic disorder produced faster response times to
self-evaluative panic-relevant associations on an Implicit Association Test, which reflects
involuntary processing of stimuli congruent with underlying threat schemas.
Overall there is strong empirical support for the schema vulnerability hypothesis
from the information-processing literature. Findings of an automatic threat-processing
bias in nonpanic states are consistent with our contention of a prepotent, enduring schematic threat organization in panic disorder. However, it is still unclear whether the schematic content in panic disorder is highly specific to physiological and mental sensations
or more reflective of general threat themes, and whether activation of these schemas is
responsible for the catastrophic interpretation of bodily sensations. We also await the
development of a more specific panic belief questionnaire that can test the predictive
validity of the schema vulnerability hypothesis in prospective research designs.
300
TREATMENT OF SPECIFIC ANXIETY DISORDERS
Hypothesis 3. Catastrophic Misinterpretation
Panic attacks are characterized by a misinterpretation of bodily or mental sensations
as signifying an imminent physical, mental, or social catastrophe. Production of the
catastrophic misinterpretation will increase panic symptoms in individuals with panic
disorder, whereas correction of the misinterpretation will prevent panic attacks (see D. M.
Clark, 1996).
Over the years various reviews of the relevant literature have concluded that there is
strong support that individuals with panic disorder are significantly more likely to misinterpret bodily sensations in terms of a serious impending threat or danger than nonpanic comparison groups (for reviews, see Austin & Richards, 2001; Casey et al., 2004;
Khawaja & Oei, 1998). Moreover, there is considerable empirical evidence that panic
disorder is characterized by elevated scores on the ASI Fear of Somatic Sensations subscale, a finding entirely predicted by the catastrophic misinterpretation hypothesis (e.g.,
Deacon & Abramowitz, 2006a; Rector et al., 2007; Taylor, Zvolensky, et al., 2007; see
also discussion of anxiety sensitivity in Chapter 4). And yet, dissenting views have been
expressed stating that a number of key aspects of the catastrophic misinterpretation
hypothesis remain in doubt (McNally, 1994; Roth et al., 2005). Three types of research
provide a critical test of the catastrophic misinterpretation hypothesis: self-report measures of catastrophic cognitions, clinical studies of the relation between misinterpretations of bodily sensations and subsequent panic symptomatology, and evidence of cognitive mediation in biological challenge experiments.
Various clinical studies indicate that most individuals with panic disorder report
thoughts or images of physical or mental catastrophe in response to internal stimuli
during panic episodes (e.g., Argyle, 1988; Beck et al., 1974; Ottaviani & Beck, 1987).
The Agoraphobic Cognitions Questionnaire (ACQ) assesses the frequency of maladaptive thoughts about catastrophic consequences (e.g., fainting, choking, heart attack,
loss of self- control) when feeling anxious (Chambless, Caputo, Bright, & Gallagher,
1984). Individuals with panic disorder score significantly higher than depressed and
other anxiety disorder groups on the ACQ Physical Concerns but not the Social/Behavioral Consequences factor (Chambless & Gracely, 1989).
D. M. Clark et al. (1997) developed the Body Sensations Interpretation Questionnaire (BBSIQ) to assess endorsement rates and belief in threatening, positive, or neutral
explanations for ambiguous panic body sensations and external events (control items).
Analysis revealed that panic disorder patients ranked negative interpretations of panic
body sensations as significantly more probable and believed the negative explanation
more than GAD, social phobia, or nonclinical comparison groups. Furthermore, the
BBSIQ correlated .49 with the ACQ Physical Concerns subscale (for similar findings, see
Austin, Richards, & Klein, 2006; Teachman et al., 2007). However, Austin et al. (2006)
found that panic patients rarely made a subsequent harm-related interpretation (e.g.,
“I’m having a heart attack”) to their initial anxiety-related interpretation (e.g., “I’m
having a panic attack”). Studies that examined interpretations to ambiguous scenarios
also found evidence of a threat interpretation bias for physical sensations in panic disorder compared to nonclinical controls (Kamieniecki et al., 1997; McNally & Foa, 1987;
see also Uren et al., 2004), although it appears the panic disorder individuals generated
more anxiety interpretations for both internal and external threats. Generally the self-
Panic Disorder
301
report studies have supported the catastrophic misinterpretation hypothesis of bodily
sensations, although most found that the interpretation bias is not specific to internal
sensations alone and that anxiety interpretations (i.e., an expectation of becoming more
anxious) are much more common than truly harm-related catastrophes (i.e., appraisals
of dying from suffocation or a heart attack).
A few studies have investigated the presence of catastrophic misinterpretations in
panic disorder samples that have been exposed to fear situations. Occurrence of a panic
attack leads to greater expectation of subsequent fear or a heightening of anticipatory
anxiety, which increases the likelihood that individuals will consider their anxious
symptoms highly threatening (i.e., Rachman & Levitt, 1985). Moreover, when panic
occurs during exposure to a fear situation, panic disorder individuals experience more
bodily sensations and catastrophic cognitions than during the nonpanic exposure trials,
although 27% (n = 8/30) of the panic episodes were not associated with any fearful cognitions (Rachman, Lopatka, & Levitt, 1988). In a further analysis of these data, Rachman, Levitt, and Lopatka (1987) found that individuals with panic disorder were four
times more likely to have a panic attack when the bodily sensation was accompanied
by catastrophic cognitions. Street et al. (1989) also found a high rate of catastrophic
thinking when individuals recorded their next three panic attacks, especially when the
attacks were expected. In addition there were many moderate correlations between the
expected disturbing cognitions and their corresponding physical sensations (see Rachman et al., 1987, for similar finding).
Kenardy and Taylor (1999) had 10 women with panic disorder use a computer diary
to self-monitor onset of panic attacks over a 7-day period. Analysis revealed that individuals overpredicted panic attacks; in 70% of cases the expectation of an attack never
materialized. Moreover, catastrophic cognitions and somatic symptoms were common
before expected but not unexpected panic attacks, indicating that catastrophic thoughts
were associated with prediction or expectation of a panic attack rather than its actual
occurrence. Finally, a small pilot study of panic disorder found that 3.25 hours of belief
disconfirmation exposure resulted in significantly greater improvement in frequency
and belief of agoraphobic cognitions as well as symptom measures than the group who
received habituation exposure training only (Salkovskis, Hackmann, Wells, Gelder, &
Clark, 2006). This suggests that reductions in catastrophic interpretations lead to an
improvement in anxious and panic symptoms. Overall, these studies support the catastrophic misinterpretation hypothesis with two caveats. First, Rachman et al. (1987) did
find a small number of “noncognitive panic attacks” that are difficult to explain from
the catastrophic misinterpretation perspective. And second, some of the expected combinations of bodily sensations and catastrophic cognitions were not found such as heart
palpitations, fear of heart attack, and various combinations of symptoms could lead to
the same catastrophic cognition.
The strongest evidence for the catastrophic misinterpretation hypothesis comes
from experiments involving panic induction via biological challenge (e.g., lactate fusion,
CO2 enriched air, hyperventilation, or exercise). There is considerable evidence that
some form of cognitive mediation is a critical factor that influences the frequency of
panic induction and heightened anxiety produced by these biological challenge experiments (D. M. Clark, 1993). In order to separate the effects of the induction and individuals’ cognitions, participants typically are randomly assigned to receive instructions
to expect that the induction would lead to unpleasant reactions or that the induction
302
TREATMENT OF SPECIFIC ANXIETY DISORDERS
would be a pleasant or benign experience. Findings from these studies indicate that type
of information provided, expectations, perceived control, and presence of safety cues
influence individuals’ anxiety and arousal to the induction (Khawaja & Oei, 1998). For
example, in a study of healthy individuals who received sodium lactate and a placebo
on two different days, only those who received the lactate infusion and anxious instructions experienced a significant increase in anxiety (van der Molen, van den Hout, Vroemen, Lousberg, & Griez, 1986).
Over the years numerous studies have demonstrated that individuals with panic
disorder show greater reactivity to inhaled carbon dioxide (CO2) than other anxiety
disorder groups and healthy controls by experiencing more intense bodily sensations,
and greater likelihood of panic symptoms and elevated anxiety as indicated by subjective measures, even though there are few differences in physiological functioning
(e.g., Perna, Barbini, Cocchi, Bertani, & Gasperini, 1995; Perna et al., 2004; Rapee et
al., 1992; Verburg, Griez, Meijer, & Pols, 1995). Furthermore, individuals with panic
disorder report that the symptoms produced by CO2 inhalation are similar to real-life
panic attacks (Fyer et al., 1987; van den Hout & Griez, 1984; see review by Rapee,
1995a). It appears that affective response to CO2 inhalation may even have etiological
significance. In a 2-year follow-up study, Schmidt, Maner, and Zvolensky (2007) found
that CO2 reactivity predicted the later development of panic attacks. However, there are
individual differences even among panic disorder individuals in their response to CO2
inhalation, with 55–80% reporting a panic attack (Perna et al., 1995, 2004; Rapee et
al., 1992). Rapee (1995a) noted that individuals who respond to a biological challenge
are more likely to experience symptoms that are similar to their real-life panic symptoms and to report thoughts of impending catastrophe. He concluded that individuals
will exhibit a greater affective response to biological challenges if they associate an
immediate impending physical or mental catastrophe (threat) with the induced sensations and perceive diminished control over the aversive experience (e.g., Rapee et al.,
1992; Sanderson, Rapee, & Barlow, 1989). Consistent with this conclusion Rapee et al.
(1992) found that the only significant predictor of fear associated with hyperventilation
and CO2 inhalation was ASI Total Score (see also Rassovsky et al., 2000). Overall these
findings are entirely consistent with the catastrophic misinterpretation hypothesis.
Additional support for the hypothesis is evident in recent studies that investigated
information processing of physical stimuli and induction of physical symptoms. Using a
modified semantic priming experiment, Schneider and Schulte (2007) found that individuals with panic disorder exhibited a significantly higher automatic (but not strategic)
priming effect for idiographically selected anxiety symptom primes followed by catastrophic interpretations than nonclinical controls. The authors interpret this automatic
priming effect as a consequence of strong idiographic associations produced by the relation of catastrophic thoughts to bodily symptoms during panic attacks. More specifically, there is evidence that imposing a respiratory load influences processing bias for
negative physical words in those with fear of suffocation (Kroeze et al., 2005; see also
Nay, Thorpe, Robertson-Nay, Hecker, & Sigmon, 2004).
In summary there is strong empirical support for the catastrophic misinterpretation
hypothesis (see Austin & Richards, 2001; Khawaja & Oei, 1998; Casey et al., 2004;
Rapee, 1995a). Misinterpreting physical or mental sensations as signifying an imminent
threat has been consistently found in self-report, clinical, and experimental studies and
its presence influences the intensity of panic symptoms. However, there are a number of
Panic Disorder
303
issues that remain unresolved. First, there is evidence that catastrophic misinterpretations of bodily sensations may not be necessary to experience a panic attack, a finding
that directly challenges a major tenet of the catastrophic cognition model of D. M.
Clark (1988). (For further discussion of this criticism, see Hofmann, 2004a; McNally,
1994; Rachman, 2004; Roth et al., 2005.) Second, there is considerable evidence that
catastrophic misinterpretations are not sufficient in themselves to produce panic. Rapee
(1995a) has argued that perceived uncontrollability is an important cognitive variable in
panic symptoms and Casey et al. (2004) have proposed an integrated model in which the
ongoing occurrence of panic is influenced by catastrophic misinterpretations of bodily
sensations and panic self- efficacy (i.e., positive cognitions that emphasize control or
coping). We would argue that a more comprehensive cognitive model of panic is needed
(see Figure 8.1) in which the extent of dissociation between an automatic catastrophic
misinterpretation and a more realistic, benign interpretation of bodily sensations will
determine the occurrence of panic attacks (Beck, 1988). In other words, the persistence
of panic symptoms may not only depend on the occurrence of catastrophic misinterpretations but also on the inability to self- correct with a more realistic explanation of the
physical changes at the elaborative stage.
Two other criticisms of the catastrophic misinterpretation model must be mentioned. Defining what is meant by “catastrophe” has proven difficult. If a narrow definition is adopted in which catastrophe means an “imminent physical or psychological
harm” (e.g., heart attack, fainting, suffocation), than these types of interpretations are
relatively infrequent in panic disorder. Instead the most common threat interpretations
associated with physical symptoms is “fear of losing control” or “fear of an impending
panic attack,” or even some social threat such as being embarrassed in front of others
(Austin & Richards, 2001). Austin and Richards suggest that a much broader range of
outcomes should be included as “catastrophes.” Finally, more research is needed on the
causal links between body sensations, catastrophic cognitions, and panic symptoms.
Rachman (2004) has argued that it is difficult to determine if catastrophic cognitions
are the cause, the consequence, or merely a correlate of panic, although the biological
challenge experiments have been most informative in this regard.
Hypothesis 4. Interoceptive Amplification
The production of a catastrophic misinterpretation of internal cues will heighten the
intensity of the feared sensations in panic but not in nonpanic states.
According to the cognitive model, a positive feedback loop occurs with the automatic catastrophic misinterpretation of bodily sensations directly contributing to a further intensification of the physical or mental changes that were the initial source of
threat schema activation. An escalation in the feared sensations will fuel continued
activation of the physiological threat schemas, ensuring that the individual with panic
disorder becomes fixated on the catastrophic misinterpretation (Beck, 1988).
Few studies have directly investigated this hypothesis. Evidence of a moderate positive correlation between catastrophic cognitions and their corresponding bodily sensation (i.e., breathless-fear of suffocation) is consistent with the interoceptive amplification hypothesis (e.g., Rachman et al., 1987; Street et al., 1989). D. M. Clark et al. (1988)
commented on a study conducted in their laboratory in which panic patients but not
304
TREATMENT OF SPECIFIC ANXIETY DISORDERS
recovered patients or healthy controls experienced a panic attack after reading word
pair associates consisting of bodily sensations and catastrophes (e.g., palpitations-dying;
nausea-numbness). In their cardiac monitoring study of panic disorder, Pauli et al. (1991)
found that anxiety elicited by cardiac perceptions led to an increase in patients’ heart
rate during the period immediately after the cardiac perception. In another study involving a panic disorder sample, scores on the ASI Physical subscales predicted subjective
fear during a hyperventilation challenge (Brown et al., 2003). Although these studies
provide only indirect support, there is sufficient evidence to encourage further research
that bodily sensations are experienced more intensely after catastrophic misinterpretations.
Hypothesis 5. Dissociation
Individuals with panic disorder will exhibit diminished ability to employ higher order
reflective thinking to generate more realistic and benign interpretations of their fearful
internal sensations compared with individuals without panic disorder.
A critical difference between a catastrophic misinterpretation model of panic and
the cognitive model of panic proposed by Beck (1988) is the central role that dissociation of higher order reflective thinking plays in the pathogenesis of the anxiety attack.
Beck stated: “The next state which is crucial to the experience of panic, as contrasted
to simple severe anxiety, is the loss of the capacity to appraise the symptoms realistically, which is associated with the fixation on the symptoms” (1988, p. 94). Thus panic
attacks occur because the individual with panic disorder is unable to retrieve a more
realistic explanation for the sensations that counters the catastrophic misinterpretation.
Unfortunately this aspect of the cognitive model has generated little research attention
as most of the focus has been on the role of catastrophic misinterpretations of bodily
sensations.
In a questionnaire study comparing individuals with panic disorder and nonclinical
groups, Kamieniecki et al. (1997) found that individuals with panic disorder provided
significantly more anxious interpretations of ambiguous internal scenarios which were
not followed by benign alternative explanations for the elevated physical sensations
described in the scenario. The authors conclude that the panic disorder patients were
unable to reinterpret their physical state in an innocuous manner. Wenzel et al. (2005)
reported that individuals successfully treated for panic disorder scored higher on items
that reflected an ability to reason about and evaluate their anxious thoughts and symptoms more realistically than individuals who still experienced difficulties with panic.
There is also evidence that providing a more benign explanation for experimentally
induced physical sensations or safety information can reduce anxiety and increase a feeling of safety (Rachman & Levitt, 1985; Rachman, Levitt, & Lopatka, 1988; Schmidt,
Richey, Wollaway-Bickel, & Maner, 2006). If a feeling of safety is a critical factor in
the offset or termination of a panic episode (Lohr, Olatunji, & Sawchuk, 2007; Rapee,
1995a), then generating a corrective interpretation of physical arousal might be an
important factor in engendering safe feelings. The inability to self- correct catastrophic
misinterpretations would be a major obstacle to acquiring safe feelings. At the very least
these preliminary findings suggest that further investigation of the dissociation of auto-
Panic Disorder
305
matic physical threat appraisals and more realistic reinterpretations would be a fruitful
area of research.
Hypothesis 6. Safety Seeking
Avoidance and maladaptive safety-seeking behavior will intensify anxiety and panic
symptoms in those with panic disorder relative to nonpanic controls.
Since safety seeking is any cognitive or behavioral strategy that is intended to prevent or minimize a feared outcome, it includes escape as well as all forms of avoidance
(Salkovskis, 1988, 1996b). Any cognitive or behavioral strategy (e.g., controlled breathing, relaxation, sitting, being accompanied by a friend, distraction) that subverts access
to information that would disconfirm the catastrophic belief is considered maladaptive
and will contribute to the persistence of panic symptoms (D. M. Clark, 1999; Salkovskis, 1988). Lohr et al. (2007) argue that safety signals may reduce the immediate experience of fear but ultimately contribute to the maintenance of pathological fear.
As noted previously, there is an extensive research literature showing that safetyseeking behavior and avoidance contribute to the persistence of anxiety (see Chapter 3,
Hypotheses 2 and 10). Schmidt et al. (2006), for example, found that the provision of
safety cues undermined the effectiveness of safety information in reducing anxiety to
a CO2 inhalation challenge, whereas other studies found a strong link between safety
behaviors and catastrophic misinterpretations (e.g., Salkovskis et al., 1996). Lundh et
al. (1998) found that a recognition bias for safe faces correlated with avoidance of fear
situations, which suggests a strong link between pursuit of safety and avoidance. In a
naturalistic self-monitoring study, Radomsky, Rachman, and Hammond (2002) found
that individuals with panic disorder used a variety of safety-seeking strategies to hasten the end of a panic attack, the most common being an effort to calm down. Individuals believed that these strategies helped terminate panic as indicated by a reduction
in bodily sensations and fearful cognitions. Even though individuals believed that the
safety-seeking behaviors made them feel somewhat safe and reduced the likelihood of
another immediate panic attack, Radomsky and colleagues employed a repeated hyperventilation challenge test to show there was, in fact, no panic-safe refractory period.
Thus individuals may believe that safety seeking helps terminate a panic attack and
reduce the likelihood of an immediate recurrence even though the prophylactic effect of
safety seeking is highly unlikely.
Furthermore, there is evidence that a reduction in safety seeking can have positive
therapeutic effects on anxiety and panic symptoms (see Salkovskis et al., 1999; Salkovskis et al., 2006). And yet Rachman found that the provision of safety-signal training
increased predictions of safety, reduced expectations of fear, and inhibited panic when
individuals were exposed to their fear situations (Rachman & Levitt, 1985; Rachman,
Levitt, & Lopatka, 1988b). Similarly Milosevic and Radomsky (2008) found that snakefearful individuals had significant reductions in subjective anxiety and fearful cognitions
as well as increased approach behavior with a single 45-minute exposure session whether
or not they were allowed to rely on safety behavior during the exposure session.
In summary, research on safety seeking indicates that a distinction must be made
between safety-seeking behavior and feelings of safety. Producing a sense of safety
306
TREATMENT OF SPECIFIC ANXIETY DISORDERS
appears to be important in terminating, and possibly inhibiting, panic (Lohr et al.,
2007; Rapee, 1995a). However, there are clearly healthy and unhealthy ways to achieve
this state of safety (Schmidt et al., 2006). Helping individuals with panic disorder adopt
stronger beliefs in safety explanations for bodily sensations may be the most effective
approach in panic disorder, whereas reliance on actual safety-seeking behavior (e.g.,
distraction, avoidance) may block access to disconfirming evidence and contribute to
the persistence of panic symptoms, though this latter conclusion still requires considerable investigation in light of more recent findings that safety behavior may not be as
deleterious as once thought.
COGNITIVE ASSESSMENT AND CASE FORMULATION
Diagnosis and Symptom Measures
Assessment for panic disorder should begin with a structured diagnostic interview like
the SCID-IV (First et al., 1997) or ADIS-IV (Brown et al., 1994) given that panic attacks
per se are highly prevalent in all the anxiety disorders. The ADIS-IV is recommended
for the diagnosis of panic disorder because it has high interrater reliability for the disorder (k = .79; Brown, Di Nardo, & Barlow, 2001) and provides a wealth of information
on panic symptoms. It distinguishes between situationally cued and unexpected panic
attacks and severity ratings are obtained on all the DSM-IV symptoms for both fullblown unexpected panic attacks and limited symptom attacks. In addition information
is collected on extent of worry over future panic attacks, situational triggers, avoidance,
interoceptive sensitivities, safety signals, and negative impact associated with recurrent
panic attacks. The module on agoraphobia provides ratings on the degree of apprehension and avoidance associated with 20 situations commonly avoided in agoraphobia.
Various self-report panic symptom measures should also be administered as part of
the cognitive assessment. In Chapter 5 we reviewed evidence that the BAI (Beck & Steer,
1990) assesses the physiological symptoms of anxiety (e.g., Beck, Epstein, et al., 1988;
Hewitt & Norton, 1993), thus making it a particularly sensitive measure for panic disorder. Leyfer, Ruberg, and Woodruff-Borden (2006) calculated that a BAI Total Score
cutoff of 8 would identify 89% of individuals with panic disorder and exclude 97%
without panic disorder. The ASI is another measure that is highly relevant for panic (see
Chapter 4) given that individuals with panic disorder score significantly higher than
individuals with all other anxiety disorders. Below we briefly discuss four panic symptom measures that are especially useful when assessing panic disorder.
Agoraphobic Cognitions Questionnaire
The Agoraphobic Cognitions Questionnaire (ACQ) is a 15-item self-report questionnaire that assesses thoughts of perceived negative or threatening consequences (i.e., fear
of fear) associated with the physical symptoms of anxiety (Chambless et al., 1984).
Individuals with agoraphobia score significantly higher than those with other anxiety
disorders, especially on the ACQ—Physical Concerns subscale (Chambless & Gracely,
1889), and the instrument is sensitive to treatment effects (Chambless et al., 1984). Individuals with panic attacks report higher ACQ scores than those without panic attacks
(Craske, Rachman, & Tallman, 1986). The mean ACQ Total score for panic disorder
Panic Disorder
307
is approximately 28, with posttreatment scores dropping to 19 (e.g., D. M. Clark et al.,
1994).
Body Sensations Questionnaire
The Body Sensations Questionnaire (BSQ) is a 17-item questionnaire also developed by
Chambless et al. (1984) to assess intensity of fear associated with physical symptoms of
arousal (Antony, 2001a). The BSQ and ACQ are normally administered together and
both have been used extensively in the research literature. Individuals with agoraphobia
or panic disorder score significantly higher on the BSQ (Chambless & Gracely, 1989)
with panic disorder samples (M = 46.3; SD = 8.7) scoring significantly higher than
healthy (M = 28.4, SD = 6.5) controls (e.g., Kroeze & van den Hout, 2000b). The BSQ
also is sensitive to treatment effects, with posttreatment scores dropping within the
normal range (i.e., D. M. Clark et al., 1994). The clinician will find the ACQ useful for
assessing exaggerated threat appraisals of physical symptoms and the BSQ useful for
assessing fear of panic-relevant bodily sensations. A copy of both measures can be found
in Antony (2001a, Appendix B).
Mobility Inventory for Agoraphobia
The Mobility Inventory for Agoraphobia (MI) is a self-report questionnaire that assesses
the severity of agoraphobic avoidance, frequency of panic attacks, and size of safety
zone (Chambless, Caputo, Jasin, Gracely, & Williams, 1985). The first section of the
questionnaire lists 26 situations often avoided in agoraphobia and individuals rate the
extent of avoidance of each situation on a 5-point scale (1 = “never avoid”; 5 = “always
avoid”) when accompanied and when alone. They then circle the five situations that
cause the greatest amount of concern or impairment. The most recent version of the MI
also instructs individuals to indicate the frequency of panic attacks in the past 7 days as
well as in the past 3 weeks, and to rate the severity of their panic attacks on a 1 (“very
mild”) to 5 (“extremely severe”) scale (see Antony, 2001b). The modified MI added
a fourth section in which individuals report on the location and size of their safety
zone. Most research on the MI has focused on the first section of the questionnaire in
which two summed scores are calculated, an Avoidance Accompanied and an Avoidance Alone score.
Individuals with agoraphobia score significantly higher on the MI Avoidance Alone
and Avoidance Accompanied subscales than do those with other anxiety disorders and
nonclinical controls (Chambless et al., 1985; Craske et al., 1986) and the factorial structure of the MI showed high stability over a 5-year period (Rodriguez, Pagano, & Keller,
2007). For the clinician the MI yields valuable information on the nature and extent of
agoraphobic avoidance often associated with panic disorder. The original MI was reproduced in an appendix of Chambless et al. (1985) and the modified MI can be found in
Antony (2001b, Appendix B).
Albany Panic and Phobia Questionnaire
The Albany Panic and Phobia Questionnaire (APPQ) is a 27-item questionnaire that
assesses level of fear (0–8 scale) associated with physical and social activities that pro-
308
TREATMENT OF SPECIFIC ANXIETY DISORDERS
duce somatic sensations (Rapee, Craske, & Barlow, 1994–1995). Three subscales are
derived; Social Phobia (10 items), Agoraphobia (nine items), and Interoceptive (eight
items). Although the factorial structure of the APPQ has been supported, contrary to
expectation APPQ Agoraphobia was more strongly related to fear of panic than APPQ
Interoceptive (Brown, White & Barlow, 2005). Until more is known about the APPQ’s
psychometric properties, it is recommended for research purposes only.
Clinician Guideline 8.8
A standard pretreatment assessment of panic disorder should include the ADIS-IV for diagnostic information as well as the BAI, ASI, ACQ, and BSQ to determine frequency and intensity of panic symptoms. The MI should be administered when agoraphobia is present.
Case Conceptualization
Although diagnostic and symptom measures are helpful in developing a case conceptualization, idiographic assessment of key cognitive and behavioral features of panic is
essential in making a case formulation. Table 8.6 provides a summary of the key elements in a cognitive assessment and case formulation for panic disorder.
TABLE 8.6. Key Elements of a Cognitive Assessment and Case Formulation of Panic Disorder
Cognitive construct assessed
Assessment instruments
Context and frequency of panic
ADIS-IV, Weekly Panic and Acute Anxiety Log (Appendix
8.1), Situational Analysis Form (Appendix 5.2)
Heightened sensitivity and vigilance of
bodily/mental sensations
BSQ, Physical Sensations Self-Monitoring Form (Appendix
5.3), Expanded Physical Sensations Checklist (Appendix
5.5)
Catastrophic misinterpretation(s)
ACQ, Physical Sensations Self-Monitoring Form (Appendix
5.3), Apprehensive Thoughts Self-Monitoring Form
(Appendix 5.4)
Beliefs, apprehension, and intolerance of
anxiety and discomfort
ASI, Identifying Anxious Thinking Errors (Appendix 5.6),
Worry Self-Monitoring Form A (Appendix 5.8)
Escape, avoidance, and other safetyseeking cognitive and behavioral strategies
MI, Behavioral Response to Anxiety Checklist (Appendix
5.7), Cognitive Responses to Anxiety (Appendix 5.9),
Exposure Hierarchy (Appendix 7.1)
Accessibility of reappraisal schemas
Symptom Reappraisal Form (Appendix 8.2), Weekly Panic
and Acute Anxiety Log (Appendix 8.1)
Outcome of panic attacks; sense of safety
and perceived coping ability
Weekly Panic and Acute Anxiety Log (Appendix 8.1),
Symptom Reappraisal Form (Appendix 8.2), Anxious
Reappraisal Form (Appendix 5.10)
Note. ADIS-IV, Anxiety Disorders Interview Schedule for DSM-IV; BSQ, Body Sensations Questionnaire; ACQ, Agoraphobic Cognitions Questionnaire; ASI, Anxiety Sensitivity Index; MI, Mobility Inventory.
Panic Disorder
309
Weekly Panic Log
One of the most important instruments in any assessment of panic is a daily self-reported
measure of panic attacks called the panic log (Shear & Maser, 1994). The panic log
should be introduced at first contact with the client and utilized as a weekly homework
assignment throughout the course of therapy. Appendix 8.1 provides a weekly panic log
that is tailored to the cognitive therapy discussed in this chapter. If completed correctly
it will give the clinician most of the basic information that is needed to develop a cognitive case formulation of panic. The panic log provides crucial contextual information
about panic attacks, their symptom expression, anxious interpretation, extent of reappraisal capacity, and coping resources. To maximize the clinical utility of the panic log,
the therapist should provide instructions on how to use the panic log. The following
points should be covered in the explanation.
1. Complete the log as soon as possible after experiencing a panic or anxiety attack
to ensure greater accuracy of self-observations.
2. Record a broad range of panic experiences including full-blown panic attacks,
partial attacks, and acute anxiety attacks. In the Severity/Intensity column, label
each anxiety episode as a full-blown panic attack (i.e., abrupt onset involving
four or more physical symptoms), a limited panic attack (i.e., abrupt onset involving one-to-three physical symptoms), or an acute anxiety episode (i.e., sudden
onset of apprehension or nervousness).
3. Duration of panic (column 1) is defined as the length of time the panic lasts at its
peak intensity (i.e., Brown et al., 1994).
4. In the second column briefly note the circumstances or context in which the anxiety or panic occurred. Make particular note of any specific external or internal
stimulus that may have triggered the panic (e.g., “you are driving alone in the
car and notice that you are breathing more deeply than usual”). Also indicate
whether the attack is expected or unexpected.
5. Briefly describe the physical and mental symptoms that characterized the panic
attack. Make special note of the symptoms that were particularly intense or
most distressing.
6. In the column labeled “Anxious Interpretation,” answer “What concerned you
most while having the panic attack?”, “What were you afraid might happen?”,
“When you were most anxious, what was the worst consequence or outcome
that crossed you mind? (e.g., heart attack, loss of control, embarrassment or
humiliation).”
7. The sixth column, labeled “Evidence for the Alternative,” inquires whether the
client was able to find any evidence or explanation that the panic attack was less
serious than first thought. “Was there anything about the anxiety or panic that
made you think it was not a serious threat?” “Or did you recall anything that
made you question the seriousness of the anxiety or panic experience?”
8. In the last column indicate how the panic or anxiety episode terminated. “Did
you do anything that ended the panic attack?” “How effective were you at bringing an end to the anxiety or panic episode?” “To what extent was a sense or feeling of safety restored at the termination of the episode?”
310
TREATMENT OF SPECIFIC ANXIETY DISORDERS
Important contextual and phenomenological information on panic can also be
obtained from the panic disorder module of the ADIS-IV. The Situational Analysis
Form (Appendix 5.2) is an alternative measure that can be used to gather data on the
situational triggers, primary symptoms, and anxious interpretation of panic. Whether
this form is used or the weekly panic log, arriving at a valid case formulation depends on
obtaining this “online assessment” of multiple instances of panic that occur in naturalistic settings. Individuals who refuse to fill in the panic log or who provide insufficient
information will hamper treatment.
Helen, who was introduced at the beginning of this chapter, recorded one to two
daily panic and anxiety episodes on her weekly panic log at pretreatment. Only one to
two of these weekly episodes were considered full-blown panic attacks. The remainder were limited-symptom attacks or acute anxiety over physical symptoms associated
with a heightened degree of worry that a panic attack might occur. A variety of situations were identified that triggered anxiety and panic including public settings, staying
overnight away from home, driving alone in the car outside her community, being in
locations that were distant from medical facilities, and the like. Evidence of mild to
moderate agoraphobic avoidance indicated that in vivo exposure should be a prominent
feature of the treatment plan.
Interoceptive Hypersensitivity
Two issues are particularly important when assessing hypersensitivity to bodily sensations. What is the first physical or mental sensation experienced in the sequence of sensations that leads to panic? And which physical or mental sensation is the focus of the
catastrophic misinterpretation?
Although the BSQ can be helpful in assessing responsiveness to bodily sensations,
the idiographic rating forms such as the Physical Sensation Self-Monitoring Form
(Appendix 5.3) or the Expanded Physical Sensations Checklist (Appendix 5.5) will have
the greatest clinical utility along with the weekly panic log. The cognitive therapist
should review completed forms with clients, extracting from the discussion the temporal order of the internal sensations and the primary sensation that is considered most
threatening. For example, a review of Helen’s panic logs revealed that the first sensation
she often noticed during a panic episode was a sense that maybe her breathing was a
little irregular followed by other sensations such as tension, weakness, restlessness, and
lightheadedness. This culminated very rapidly in the physical symptom that was the
focus of her catastrophic misinterpretation and the apex of the panic experience: shortness of breath. Based on this information we included symptom amplification exercises
in our treatment plan in order to increase Helen’s exposure to the breathlessness sensation and decatastrophize her interpretation of the sensations.
Catastrophic Misinterpretation
A critical part of the cognitive assessment is to identify the primary catastrophic misinterpretation of internal sensations. The clinician focuses on discovering the impending
immediate physical or mental catastrophe that underlies the panic episode (e.g., fear
of heart attack, suffocation, going crazy). Often a fear of anxiety or dread of future
panic attacks replaces the somatic catastrophe for those with a history of recurrent
Panic Disorder
311
panic attacks. For others, fear of panic, loss of control, and intolerance of anxiety are
associated features of the catastrophic misinterpretation. Although Helen’s catastrophic
misinterpretation remained fear of suffocation, in later sessions she expressed greater
anxiety and apprehension about the return of panic attacks rather than of dying from
suffocation. In the early stage of treatment it is important to obtain a full description
of the various negative consequences that clients think about when they are anxious or
panicky. Helen’s treatment plan required that we target both her catastrophic misinterpretation of chest pain and breathlessness (i.e., fear of heart attack or suffocation) and
her apprehension about panic and intolerance of anxiety.
As noted in Table 8.6, the ACQ can provide some initial indication of the patient’s
misinterpretation of anxious symptoms. However, self-monitoring forms that instruct
individuals to record their symptom appraisals during peak anxiety will be most helpful. These include the weekly panic log, the Physical Sensations Self-Monitoring Form
(Appendix 5.3), and the Apprehensive Thoughts Self-Monitoring Form (Appendix 5.4).
It may be necessary to use a panic induction exercise during the session to identify the
client’s faulty appraisal process. This may be especially true for individuals who have
limited insight into their anxious cognitions.
Apprehension and Intolerance of Anxiety
It is important to identify the panic individual’s faulty cognitions and beliefs about anxiety, panic, and physical discomfort more generally. The ASI will provide an indication
of an individual’s tolerance of anxiety, especially its physical symptoms. Faulty beliefs
about anxiety can also be deduced from the types of cognitive errors that individuals
commit when anxious (use Identifying Anxious Thinking Errors, Appendix 5.6) and
the focus of their worries (use Worry Self-Monitoring Form, Appendix 5.8). Individuals
with panic disorder often worry about being anxious and panicky, so their worry content may reveal their beliefs about anxiety and its consequences. Helen had a very good
response to cognitive therapy for panic but continued to endorse a number of beliefs that
ensured recurrent states of heightened anxiety such as “If I have some unexpected physical discomfort, there must be something wrong,” “I have to deal with this discomfort,
or it could escalate into anxiety and panic,” “I can’t stand feeling anxious, I have to get
rid of the feeling,” and “If I don’t stop the anxiety, it will escalate into panic.” Thus the
latter sessions shifted focus from the catastrophic misinterpretation to normalization
exercises designed to increase her tolerance of anxiety.
Avoidance and Safety-Seeking
A cognitive assessment of panic must also include a list of all the situations and stimuli,
both external and internal, that are avoided for fear of elevated anxiety or panic. For each
situation the patient should rate degree of anxiety associated with the situation (0–100)
and extent of avoidance (0= never avoided to 100= always avoided). In addition the cognitive therapist identifies all the subtle cognitive and behavioral safety cues that may be
used to reduce anxiety. The Behavioral Responses to Anxiety Checklist (Appendix 5.7)
and the Cognitive Responses to Anxiety Checklist (Appendix 5.9) forms can be helpful
in this regards, whereas the Mobility Inventory and Exposure Hierarchy (Appendix 7.1)
may be used to explore avoidance behavior. If the concept of avoidance is broadened to
312
TREATMENT OF SPECIFIC ANXIETY DISORDERS
include interoceptive and experiential states (i.e., White et al., 2006), then the clinician
should take a wide-ranging perspective when describing the avoidance component of the
case formulation. As noted previously, Helen continued to use avoidance to manage her
anxiety so in vivo exposure was a critical component of her treatment plan.
Reappraisal Capacity
In the current cognitive model of panic disorder, loss of reappraisal capability is an
important factor in the persistence of panic attacks. Therefore it is important to assess
an individual’s ability to generate alternative, nonthreatening explanations for her physical sensations. The Symptom Reappraisal Form (Appendix 8.2) can be used to assess
critical components of reappraisal capability that might be present prior to treatment.
Three particular questions need to be addressed.
1. Is the client able to offer a number of alternative non-threat explanations for the
physical sensations?
2. How much does he believe these explanations when anxious or panicky and
when not anxious?
3. Is the client able to recall these explanations when anxious and if so, what effect
does this have on the anxious state?
The weekly panic log can be a useful starting point for a discussion on possible alternative explanations for unpleasant or anxious physical sensations. Even if an individual is
unable to generate an alternative explanation to the catastrophic misinterpretation, this
will be valuable clinical information for treatment planning.
In our case illustration, Helen’s initial apprehensive thoughts after noticing an unexpected physical sensation were “What’s wrong with me?”, “Why am I feeling this way?”
She immediately generated a catastrophic misinterpretation such as “Could this be a
heart attack?” (i.e., if she felt chest pain), “What if I can’t catch my breath and then start
to suffocate?” (i.e., if she experienced a breathless sensation), or “Will I have a terrible
panic attack?” At pretreatment she was able to generate two less-threatening alternative explanations for the sensations (e.g., the sensation could be a symptom of anxiety
or stress that will eventually subside). Occasionally she could attribute the symptoms
to physical activity or a state of ill health (e.g., having a cold, flu symptoms). However,
she had difficulty believing these alternative explanations or even being able to access
them when she felt intense anxiety or panic. Also she became intolerant of anxiety, so
interpreting the sensations as symptoms of anxiety provided no relief for her. It was
clear from the assessment that strengthening her reappraisal capability would be an
important focus of treatment.
Perceived Panic Outcome
A final component of the case conceptualization is to determine the “natural” outcome of
panic attacks. It is expected that individuals will engage in escape, avoidance, and safetyseeking behaviors in an effort to control the anxiety and panic. The clinician should
assess the perceived effectiveness of these strategies. To what extent is an individual able
to achieve a sense of safety after the occurrence of an anxiety or panic episode? How long
Panic Disorder
313
does this sense of safety last before the patient is again concerned about the recurrence
of panic? What is the individual’s degree of self-efficacy in her ability to cope with panic?
Information on panic outcome can be obtained from the weekly panic log, the Symptom
Reappraisal Form (Appendix 8.2), and the Anxious Reappraisal Form (Appendix 5.10).
Helen was able to achieve a reasonably high level of safety after her episodes of
acute anxiety and panic but these tended to be relatively short-lived (e.g., 12–24 hours).
She engaged in extensive reassurance seeking from family members and searching for
her symptoms on the Internet, as well as avoiding perceived triggers. She believed that
avoidance was quite effective in curbing the anxiety and ensuring that it did not escalate
into panic. The reassurance seeking was considered moderately effective in reducing
current states of anxiety over unexplained physical sensations. Helen also relied heavily
on self-reassurance in which she repeated to herself “Everything will be okay, nothing
is wrong with me.” Again she thought this helped “calm her down” to a certain extent.
Treatment, then, had to target Helen’s beliefs about the effectiveness of her avoidance
and safety-seeking strategies to ensure the elimination of maladaptive coping that contributed to the persistence of panic.
Clinician Guideline 8.9
A cognitive case formulation of panic should include a contextual analysis of the panic attacks
as well as an assessment of (1) physiological hypervigilance, (2) catastrophic misinterpretation of bodily sensations, (3) presence of maladaptive beliefs about anxiety tolerance, (4) role
of avoidance and safety-seeking strategies, (5) accessibility of reappraisal schemas, and (6)
perceived outcome of anxiety and panic episodes. The case formulation will be the basis of
treatment planning and implementation of an individualized cognitive intervention.
DESCRIPTION OF COGNITIVE THERAPY FOR PANIC DISORDER
There are five main treatment goals in cognitive therapy for panic disorder. The first
two goals pertain to the automatic schematic threat processing that occurs during the
immediate fear response (Phase I), whereas the remaining goals refer to responses that
occur during elaborative processing (Phase II) (see Figure 2.1). The primary treatment
goals are:
1. Reduce sensitivity or responsiveness to panic-relevant physical or mental sensations
2. Weaken the catastrophic misinterpretation and underlying hypervalent threat
schemas of bodily or mental states
3. Enhance cognitive reappraisal capabilities that result in adoption of a more
benign and realistic alternative explanation for distressing symptoms
4. Eliminate avoidance and other maladaptive safety-seeking behaviors
5. Increase tolerance for anxiety or discomfort and reestablish a sense of safety
Table 8.7 presents the main treatment components of cognitive therapy employed to
achieve these goals.
314
TREATMENT OF SPECIFIC ANXIETY DISORDERS
TABLE 8.7. Main Treatment Components of Cognitive
Therapy for Panic
•฀ Education into the cognitive therapy model of panic
•฀ Schematic activation and symptom induction
•฀ Cognitive restructuring of catastrophic misinterpretation
•฀ Empirical hypothesis testing of alternative explanation
•฀ Graded in vivo exposure
•฀ Symptom tolerance and safety reinterpretation
•฀ Relapse prevention
•฀ Breathing retraining (optional)
Educating Clients into the Cognitive Therapy Model of Panic
The first treatment session focuses on educating the client into the cognitive explanation for recurrent panic attacks. If the cognitive assessment strategy has been followed,
then the therapist already has much of the critical information available for educating the client such as the situational triggers for panic, distressing physical sensations,
catastrophic misinterpretations, and maladaptive avoidance/safety-seeking responses.
Normally clients have started keeping a weekly panic log (see Appendix 8.1) and so
a typical panic episode can be selected from the log. Using Socratic questioning, the
cognitive therapist explores the client’s experience during this panic episode and his
interpretation of the symptoms. The therapist and client collaboratively complete the
Vicious Cycle of Panic form found in Appendix 8.3. It is important that the therapist
records specific thoughts and feelings associated with the panic episode and that the
cognitive explanation is presented as “one possible explanation of the origins of panic
that needs to be tested.”
At this initial stage of treatment it is unlikely the client is ready to abandon her
catastrophic misinterpretation and embrace the cognitive explanation. Instead the goal
of the educational session is to merely introduce an alternative explanation for panic
that provides a treatment rationale. The session normally ends with a homework assignment in which clients continue with their panic logs but this time they examine whether
their anxiety and panic experiences are consistent or not with the cognitive explanation.
When reviewing homework in the subsequent session, it is important that the therapist
deal with anxiety experiences that appear contrary to the model and reinforce the client’s observations that are consistent with the cognitive explanation.
In our case illustration a Vicious Cycle of Panic form (see Appendix 8.3) was completed at the outset of cognitive therapy. Helen identified a number of triggers from her
panic log such as being at a work meeting and sitting beside the guest speaker, not being
in close proximity to a hospital, flying, and driving alone some distance from home.
Her initial physical sensations were feeling lightheaded, sensing that her breathing was
a little irregular, and experiencing an unusual feeling of pressure in her chest. This was
followed by some initial anxious cognitions such as “What’s wrong with me?”, “Why
am I feeling this way?”, “Something is not right,” “I don’t like this,” “I am beginning to
feel anxious,” “I feel trapped,” and so on. These anxious thoughts often led to an escalation in a few physical sensations such as feelings of suffocation or heart palpitations.
Once these intense physical sensations occurred, Helen identified a number of cata-
Panic Disorder
315
strophic cognitions like “I’m not getting enough air, I’m going to die of suffocation,”
“What if I’m having a heart attack?”, or “If I don’t stop I’m going to have a full-blown
panic attack.” The catastrophic misinterpretation led to various control efforts such as
escape, reassurance seeking from others, controlled breathing, or distraction, which
together often ended in intense anxiety or panic attacks. After completing the Vicious
Cycle of Panic form, the therapist emphasized that catastrophic misinterpretations and
maladaptive control efforts were the main catalysts for panic rather than the real possibility of some imminent threat (e.g., possible heart attack). Helen was given a copy of
the completed Vicious Cycle of Panic form and asked to record her anxiety and panic
experiences over the next week with particular focus on whether the cognitive model
was a good explanation for her anxious experiences.
Clinician Guideline 8.10
Use the Vicious Cycle of Panic form (Appendix 8.3) to begin educating clients to the cognitive model and highlight the central role of catastrophic misinterpretations in the persistence
of panic.
Schema Activation and Symptom Induction
A critical feature of cognitive therapy for panic is the use of within-session exercises to
induce the client’s feared physical sensations (Beck, 1988; Beck & Greenberg, 1988; D.
M. Clark, 1997; D. M. Clark & Salkovskis, 1986). When cognitive therapy of panic was
first developed, patients were always given a 2-minute breathing hyperventilation exercise followed by instruction in controlled breathing in order to introduce overbreathing
as a possible alternative explanation for the occurrence of intense physical sensations (D.
M. Clark & Salkovskis, 1986). However, it is now known that hyperventilation probably plays a less prominent role in panic, so controlled breathing is no longer recommended in most cases of panic disorder (see discussion below). Furthermore, cognitive
therapists are more likely to use a variety of induction exercises repeatedly throughout
treatment based on the positive effects of interoceptive exposure on panic reduction (see
White & Barlow, 2002).
Symptom induction exercises are important in cognitive therapy of panic disorder
because they allow direct activation of threat schemas and opportunity to challenge
catastrophic misinterpretations of bodily sensations. Usually the intentional production
of symptoms like dizziness, heart palpitations, breathlessness, and so on in the presence
of the therapist is less intense and better tolerated by the patient than in real life. In this
way the client learns that certain physical sensations are not always frightening, that the
physical sensations do not lead to the catastrophic outcome, and that an exacerbation
of unwanted sensations can be due to other, more benign causes. Often the withinsession symptom induction is the first direct experiential evidence that challenges the
catastrophic misinterpretation. After engaging in symptom induction, the cognitive
therapist always reviews the experience with clients in terms of whether the experience confirms or disconfirms the catastrophic misinterpretation of bodily sensations.
Symptom induction exercises are introduced by the second or third session and they
are repeated often throughout treatment. Eventually symptom induction is assigned as
316
TREATMENT OF SPECIFIC ANXIETY DISORDERS
homework with clients instructed to practice intentionally producing their feared physical sensations first in neutral and then in anxiety-provoking situations.
Before introducing symptom induction it is important to determine if the client has
any medical contraindications for engaging in the exercise. Of course clients must be
physically able to do the exercise and willing to endure a moderate level of discomfort.
Any medical problems that could be worsened by an induction exercise must be taken
into account by possibly consulting with the client’s family physician. Taylor (2006)
lists various health conditions that would warrant extreme caution when using certain
induction exercises (e.g., lower back pain, pregnancy, postural hypotension, chronic
obstructive lung disease, severe asthma, or cardiac disease).
Table 8.8 presents a list of the most common symptom induction exercises, the
physical sensations evoked by the exercise, and an example of a typical threat misinterpretation associated with the symptom. See also Taylor (2000, 2006) and Antony,
Rowa, Liss, Swallow, and Swinson (2005) for a similar list of symptom induction and
exposure exercises.
As can be seen from this table, most of these exercises are very brief and must be
repeated frequently both as within-session demonstrations and as homework assignments. Antony et al. (2005) found that breathlessness/smothering sensations, dizziness
or feeling faint, and pounding/racing heart were the most common physical sensations
elicited by the exercises. Although two-thirds of the panic disorder group in their study
reported at least moderate fear to one or more of the symptom induction exercises, most
exercises produced only a low intensity of symptoms with spinning, hyperventilation,
breathing through a straw, and use of a tongue depressor the most potent exercises.
Other exercises such as quickly raising the head, staring at a light, tensing muscles, running on the spot, or sitting close to a heater were relatively ineffective.
Hyperventilation and breath holding were the two main symptom induction exercises used with Helen. These proved highly effective because of her fear of suffocation.
Breath holding, in which Helen was encouraged to hold her breath until she felt absolutely compelled to breathe, was a particularly effective intervention that was first demonstrated in session and then assigned whenever she felt anxious about her breathing.
By holding her breath, Helen was challenging her catastrophic view “I can’t breathe”
and by exaggerating the sense of breathlessness the sensation became less frightening.
The intense urge to breathe after a period of holding her breath was powerful evidence
that “not breathing” was extremely difficult to do even when it was intentional. Her
panicogenic belief that “I might just stop breathing and die” was weakened by realizing
that she possessed an intense automatic physiological urge to breathe.
Clinician Guideline 8.11
Within-session symptom induction is a critical therapeutic ingredient for activating panicrelevant fear schemas and directly challenging the catastrophic misinterpretation of physical sensations. A solid rationale for symptom induction must be provided. The exercises
are utilized repeatedly throughout treatment and eventually assigned as homework. Some
exercises are more effective than others in provoking physical sensations that are somewhat
similar to naturally occurring panic attacks.
Panic Disorder
317
TABLE 8.8. Symptom Induction Exercises Commonly Used in the Treatment of Panic Disorder
Exercise
Evoked physical sensation
Example of misinterpretation of threat
1. Hyperventilate for 1 minute
Breathlessness,
smothering sensation
“I can’t stand this; I think I am going to
faint if I continue.”
2. Hold breath for 30 seconds
Breathlessness,
smothering sensation
“What if I can’t breathe normally? I
could suffocate.”
3. Breathe through narrow
straw for 2 minutes
Breathlessness,
smothering sensation
“I need to get more air or I’ll
suffocate.”
4. Spin around at medium pace
while standing for 1 minute
Dizzy or faint
“If I let myself feel nauseous, I might
vomit.”
5. Place head between knees
for 30 seconds and then raise
head quickly
Dizzy or faint
“When I feel lightheaded, could this be
a sign of a stroke?”
6. Shake head rapidly from side
to side for 30 seconds
Dizzy or faint
“When feeling dizzy I am losing
contact with reality which could lead to
insanity.”
7. Tense all body muscles for 1
minute
Trembling, shaking
“People will notice that I am trembling
and think there is something wrong
with me.”
8. Run on spot for 1 minute
Pounding, racing heart
“I could have a heart attack.”
9. Sit facing a heater for 2
minutes
Breathless, smothering
sensation, sweating
“People will be disgusted by my
sweating.”
10. Place tongue depressor at
back of tongue for 30 seconds
Choking sensation
“This choking feels so bad it could
cause me to vomit.”
11. Stare continuously at
fluorescent light for 1 minute
and then try to read
Dizzy or faint; feeling
of unreality
“My environment is feeling weird. This
could mean that I am starting to go
insane.”
12. Stare continuously at self in
mirror for 2 minutes
Feeling unreal, dreamy;
dizzy or faint
“If I feel spacey I could lose contact
with reality.”
13. Stare continuously at spot on
wall for 3 minutes
Feeling unreal, dreamy;
dizzy or faint
“Feelings of unreality means that I
could be having a stroke.”
Cognitive Restructuring of Catastrophic Misinterpretation
Cognitive restructuring fulfills two functions in cognitive therapy of panic: it introduces
conflicting evidence for the catastrophic misinterpretation and it offers an alternative
explanation for internal sensations. In panic disorder evidence gathering, identifying
cognitive errors (i.e., exaggerating the probability and severity of imminent danger), and
generating alternative explanations will be most helpful. See Chapter 6 for a detailed
discussion of these cognitive intervention strategies.
It is often useful to begin cognitive restructuring with a very clear description of
the most feared catastrophic outcome and then generate a list of possible alternative
explanations for the physical sensations. The Symptom Reappraisal Form (Appendix
318
TREATMENT OF SPECIFIC ANXIETY DISORDERS
8.2) can be used to focus the client on alternative explanations for fearful sensations.
Most clients have considerable difficulty generating alternative explanations for their
most feared sensations so this will take a considerable amount of guided discovery. A
variety of alternative explanations for the symptoms can be raised such as (1) response
to heightened anxiety; (2) reaction to stress; (3) product of physical exertion; (4) fatigue;
(5) side effects of coffee, alcohol, or medication; (6) heightened vigilance of bodily sensations; (7) strong emotions like anger, surprise, or excitement; (8) random occurrence of
benign internal biological processes; or (9) other context-specific possibilities.
Another aspect of the alternative explanation that is emphasized is the role that catastrophic thoughts and beliefs play in exacerbating symptoms (D. M. Clark, 1996). For
example, “Is an underlying cardiac condition your problem so that chest pains could signal
a heart attack (catastrophic interpretation) or is your problem that you believe there is
something wrong with your heart and so you are preoccupied with your heart rate” (alternative cognitive explanation)? At this point the therapist simply raises these alternative
explanations as possibilities or hypotheses and invites the client to investigate the validity of each explanation by gathering confirming and disconfirming evidence. This can be
done by using information recorded on the Weekly Panic Log (Appendix 8.1) or one of the
cognitive forms provided in Chapter 6 (e.g., Appendices 6.2 or 6.4). The goal of cognitive
restructuring is for individuals with panic to realize that their anxiety and panic symptoms
are due to their erroneous beliefs that certain physical sensations are dangerous. Although
patients may find it difficult to accept this alternative because of their heightened anxiety,
they are repeatedly encouraged to focus on the evidence, not on how they feel.
A major part of Helen’s cognitive therapy for panic was the gathering of evidence
for alternative explanations for her symptoms of breathlessness, which had become the
primary dreaded physical sensation. Gradually, with accumulating evidence based on
repeated experiences, she began to accept that her sense of breathlessness was most
likely due to excessive monitoring of her breathing and the possibility that she was
actually suffocating was entirely remote at best. Over time she found evidence that
other physical sensations were probably due to stress, anxiety, fatigue, or alcohol consumption was much more compelling than the automatic catastrophic interpretation. At
this point therapy shifted away from challenging the catastrophic interpretation toward
increasing her tolerance of anxiety and its physical manifestations.
Clinician Guideline 8.12
In panic disorder cognitive restructuring focuses on gathering evidence (1) that the client
automatically generates a highly unlikely and exaggerated misinterpretation of unwanted
physical or mental sensations, and (2) that alternative, benign explanations are more plausible. The role of catastrophic thoughts and beliefs in perpetuating anxiety and panic symptoms is emphasized throughout treatment.
Empirical Hypothesis-Testing Experiments
Behavioral experiments play a particularly important role in the treatment of panic.
They often take the form of deliberate exposure to anxiety-provoking situations in
order to induce fearful symptoms and their outcome. The outcome of the experiment
Panic Disorder
319
is observed and provides a test of the catastrophic versus the alternative explanation
for bodily sensations. D. M. Clark and Salkovskis (1986) describe various behavioral
experiments that can be used in the treatment of panic disorder.
A number of behavioral experiments were used to test Helen’s catastrophic interpretations and beliefs. In one homework assignment she was asked to hold her breath
whenever she felt breathless sensations in order to amplify the sensations. After a few
seconds of breath holding, she was told to breathe normally and note differences between
breath holding and breathing. “Was there any evidence that she was exaggerating the
sense of breathlessness prior to breath holding?” “Was she able to breathe normally
after holding her breath?” From these experiences Helen found evidence that indeed
she was exaggerating breathlessness and her breathing was much more normal than
she thought. In another behavioral experiment Helen was encouraged to induce physical sensations while in fear situations by increasing her physical activity level. These
experiments provided evidence that physical sensations themselves do not automatically
lead to anxiety or panic (e.g., “Even when anxious, increasing my heart rate by running up stairs does not increase my anxiety level”). Instead she discovered that how she
interprets the symptoms determines whether anxiety escalates into panic (e.g., “When I
know my heart is pounding fast because of exercise I don’t feel anxious”).
Clinician Guideline 8.13
Behavioral experiments provide a critical test of the role that catastrophic thoughts and
beliefs play in the persistence of anxiety and panic symptoms. The experiments are designed
to show that the mere occurrence of physical sensations is not the primary cause of anxiety
but rather it is their catastrophic misinterpretation that leads to panic attacks.
Graded In Vivo Exposure
Given that most individuals with panic disorder exhibit at least mild forms of agoraphobic avoidance, graded in vivo exposure is a major component of cognitive therapy
for panic disorder. When agoraphobic avoidance is severe, in vivo exposure must be
introduced early in treatment and become the main focus of therapy. However, the cognitive therapist uses exposure to challenge the catastrophic cognitions and beliefs of the
agoraphobic individual. Since Chapter 7 provided an extensive discussion of graded in
vivo exposure and its implementation, the reader is encouraged to consult that section
when employing exposure exercises in cognitive therapy of panic.
In our case illustration Helen presented with fairly extensive avoidance of external
situations because of her fear of panic attacks and of being too distant from a hospital
in case she suffered a heart attack or episode of suffocation. A fear hierarchy was constructed involving 23 situations ranging from taking a bus trip to a nearby city (rated
10 on a scale of 0–100) to taking a transcontinental flight (rated 100). Helen engaged
in repeated exposure to a variety of situations on her fear hierarchy, gathering evidence
against her most feared outcomes, and confirming the role of catastrophic thinking
in the genesis of panic. Furthermore, the exposure suggested more benign, alternative
explanations for her physical sensations, thereby enhancing her ability to reappraise
unwanted feelings and sensations.
320
TREATMENT OF SPECIFIC ANXIETY DISORDERS
Clinician Guideline 8.14
Graded in vivo exposure is important in the treatment of agoraphobic avoidance and in disconfirming the catastrophic thoughts and beliefs pathogenic to panic attacks and their fear.
Symptom Tolerance and Safety Reinterpretation
As stated earlier, cognitive therapy seeks to increase the panic individual’s tolerance
of unexpected physical sensations and discomfort as well as subjective anxiety and to
instill a greater sense of safety and coping ability. This can be accomplished by intentionally focusing on the client’s ability to tolerate the physical symptoms of anxiety
during within-session and between-session behavioral exposure exercises. For example,
a client who is anxious about chest tightness and heart palpitations could be asked to
monitor his physical sensations while doing a cardio workout in the gym. Not only
would repeated experiences of physiological activation provide evidence that physical
symptoms can be tolerated, but the panic-prone individual will be learning that mere
occurrence of physical symptoms is not dangerous. However, for these experiences to be
therapeutic the cognitive therapist must repeatedly emphasize the idea that “clearly you
are able to tolerate these physical sensations just like everyone else.”
Therapy must also focus on increased tolerance for physical discomfort and anxiety.
Clients could be asked to keep a diary of their experiences of physical discomfort that
are not associated with anxiety such as episodes of headaches, sore muscles, fatigue, and
the like. Individuals can be asked to rate the degree of discomfort associated with these
symptoms and their level of anxiety. The point of this exercise is for the panic-prone
individual to learn that she is capable of tolerating discomfort without feeling anxious.
By reinforcing this observation, the therapist can strengthen the client’s beliefs in her
ability to cope with the physical discomfort associated with anxiety. Moreover, tolerance of anxiety can be improved through graded in vivo exposure exercises in which
the therapist gradually increases the level of anxiety so individuals learn they can cope
with even high anxiety states.
The cognitive therapist can increase the client’s sense of safety by helping him reinterpret anxiety-provoking situations encountered during homework assignments. At
every opportunity, the therapist redirects the client’s attention by asking questions such
as “What aspects of this situation suggested safety?”, “Was there anything about this
situation that made you think it was less dangerous and more safe than you initially
thought?”, or “As you look back on the situation, what safety cues were present that
you just didn’t notice at the time?” An important goal of cognitive therapy is to “train”
the individual with panic disorder to intentionally reconsider the safety features of an
anxiety-provoking situation in order to counter his automatic catastrophic interpretation. The Symptom Reappraisal Form (Appendix 8.2) can be used for this purpose. In
addition it is helpful to have clients rate the “realistic” level of danger associated with
the situation (e.g., 0–100 rating scale) as well as the “realistic” level of safety after
recording danger and safety features on the panic log (Appendix 8.1). It is important to
ensure that clients’ ratings are based on a “realistic” assessment rather than on “how
they feel” because emotion-based ratings will always be distorted because of a heightened anxiety state.
Panic Disorder
321
Given Helen’s relative youth and good physical health, she was encouraged to increase
her physical activity level and record her physiological arousal. This proved quite effective
in helping Helen realize she could tolerate chest tightness and breathless sensations, and
that these sensations could be evoked without danger. Also the breath-holding exercises
when anxious again provided evidence of tolerance and safety. Later in the therapy sessions cognitive restructuring always focused on processing the safety features of anxious
experiences. Helen was repeatedly asked questions such as “Looking back, what aspects
of the situation indicate that it was safer than you originally thought”? Toward the end
of treatment, Helen would spontaneously generate safety reinterpretations of anxietyprovoking situations and reported a greater sense of safety in her daily life.
Clinician Guideline 8.15
A perceived sense of safety and tolerance of the physical symptoms of anxiety are important goals for cognitive therapy of panic. They are achieved by cognitive restructuring and
behavioral exercises that emphasize the client’s natural tolerance of discomfort and the reinterpretation of safety features associated with anxiety-provoking situations.
Relapse Prevention
As is done in the treatment of other anxiety disorders, relapse prevention should be built
into the final sessions of cognitive therapy for panic. The therapist must ensure that the
client realizes that occasional panic attacks will occur, that unexpected physical sensations may occur from time to time, and that anxiety is a normal part of life. Relapse can
be minimized if the client has realistic expectations of treatment outcome and adopts a
healthy perspective on anxiety and panic. In addition, significant reduction in the client’s “fear of fear” can improve the chance of reduced relapse and recurrence of panic.
The client who continues to fear panic attacks (e.g., “I just hope I never have to experience those terrible panic attacks again”) is probably most vulnerable to relapse when the
physical symptoms of anxiety reoccur.
In addition to correcting unrealistic expectations about treatment and the “return
of fear,” a number of other measures can be taken to prevent relapse. Therapy sessions
can be gradually faded and booster sessions scheduled. An intervention protocol can be
written down that clearly specifies what to do if unexpected physical symptoms return
or the individual experiences a resurgence in anxiety. However, the most important
relapse prevention strategy for panic may involve having panic disorder patients intentionally produce their feared physical sensations when in anxiety-provoking situations.
Those individuals who progress to the point where they can exaggerate their physical
symptoms while feeling highly anxious may be better inoculated against future unexpected resurgences of anxiety and panic.
Clinician Guideline 8.16
Relapse prevention is enhanced when cognitive therapy clients are prepared for the unexpected return of fear and panic. In addition relapse and recurrence of panic disorder may
322
TREATMENT OF SPECIFIC ANXIETY DISORDERS
be less likely in individuals who can engage in exaggerated physiological activation when
experiencing high levels of anxiety.
Breathing Retraining (Optional)
Breathing retraining is a relaxation strategy that was incorporated into early versions of
cognitive therapy for panic disorder (e.g., Beck, 1988; Beck & Greenberg, 1988; D. M.
Clark, 1986a). Based on the notion that hyperventilation, which involves deep and rapid
breathing, is an important factor in the production of panic attacks, it was proposed
that training in slow, shallow breathing should counter panic symptoms (D. M. Clark,
Salkovskis, & Chalkley, 1985; Salkovskis, Jones, & Clark, 1986). Key elements of D. M.
Clark and Salkovskis’s (1986) early cognitive treatment protocol for panic included:
1. A 2-minute voluntary hyperventilation exercise of breathing at a rate of 30
breaths per minute.
2. Observation of the physical sensations caused by hyperventilation and their similarity to panic symptoms.
3. Education on the physiology of hyperventilation and how it can produce the
physical sensations of a panic attack.
4. Reattribution of the physical symptoms of panic to stress-induced hyperventilation (or overbreathing) rather than to a misperceived catastrophic health threat
(e.g., “I am having a heart attack”).
5. Training in slow breathing in order to provide a coping response that is incompatible with hyperventilation. Controlled breathing also becomes a behavioral
experiment by demonstrating that physical symptoms must be due to overbreathing rather than the catastrophic threat because the symptoms are so quickly
reduced when slow, shallow breathing is established.
The hyperventilation exercise and breathing retraining became key elements of the
cognitive therapy treatment protocol for panic offered at the Center for Cognitive Therapy in Philadelphia from the mid-1980s to the late 1990s (Beck & Greenberg, 1988).
Together they provided a critical empirical hypothesis-testing experiment indicating
that the catastrophic misinterpretation of symptoms was incorrect, and that the physical
symptoms were actually a harmless consequence of overbreathing or even hyperventilation (Beck & Greenberg, 1987).
Breathing Retraining Exercise
Abdominal or diaphragmatic breathing has been the most common form of breathing
retraining used in the treatment of anxiety disorders. It assumes a key role for hyperventilation in the etiology of panic by causing an acute decrease in arterial partial
pressure carbon dioxide (pCO2), called hypocapnia, that in turn results in a wide range
of unpleasant bodily sensations (e.g., dizziness, heart palpitations, tingling in extremities, breathlessness), which the individual misinterprets as representing a serious medical threat (Meuret, Ritz, Wilhelm, & Roth, 2005; D. M. Clark et al., 1985). Various
studies have found other breathing abnormalities in anxiety disorders such as shallow
Panic Disorder
323
and rapid overbreathing, disorganized breathing patterns, and frequent sighing (see
Meuret et al., 2005; Meuret, Wilhelm, Ritz, & Roth, 2003; Salkovskis et al., 1986).
Individuals are trained in slow, deep abdominal breathing to eliminate hypocapnia and
its uncomfortable physical sensations thereby reducing the anxious state. Table 8.9
presents a typical diaphragmatic breathing retraining protocol.
Current Status of Breathing Retraining
There is currently considerable debate over the role of breathing retraining in CBT for
panic disorder. There are three reasons why cognitive-behavioral therapists are now
questioning the use of breathing retraining. The first is a very practical, clinical concern.
Like using other forms of relaxation, a person with panic disorder might use controlled
breathing as a safety response or coping strategy to escape from an anxious state (Antony
& McCabe, 2004; Salkovskis et al., 1996; White & Barlow, 2002). This, of course, would
undermine the effectiveness of cognitive therapy by reinforcing a fear of the anxiety and
the client’s faulty evaluation of the dangerousness of the physical sensations. If there is
any evidence that the client is using controlled breathing because of a fear of anxiety and
its symptoms, then the coping response should be discontinued immediately.
Second, the rationale for offering breathing retraining in panic disorder has been
called into question with evidence that hyperventilation and hypocapnia are often not
present even in panic attacks that occur in the natural setting (see review by Meuret et
al., 2005; Taylor, 2000). And third, the therapeutic effectiveness of breathing retraining has been questioned (e.g., Salkovskis, Clark, & Hackman, 1991; D. M. Clark et
al., 1999). Schmidt and colleagues conducted a dismantling study that compared the
effectiveness of 12 sessions of group-administered CBT plus breathing retraining, CBT
without breathing retraining, and a wait list condition at posttreatment and 12-month
follow-up (Schmidt, Woolaway-Bickel, et al., 2000). At posttreatment both active treatments were significantly improved over the wait list condition but there was no statistically significant difference between the CBT and CBT + breathing retraining conditions.
At 12-month follow-up 57% of the CBT group met recovery criteria compared with
37% for the CBT + breathing retraining group. The authors concluded that the addition of diaphragmatic breathing does not add any therapeutic benefits to CBT for panic
beyond the standard treatment components of education, cognitive restructuring, and
exposure. They recommended that respiratory- control techniques be used only as a
behavioral experiment to provide corrective information for the catastrophic misinterpretation of bodily sensations and that therapists refrain from using them as an anxiety
management technique. Based on these findings we conclude that breathing retraining
should be considered optional in cognitive therapy for panic.
Clinician Guideline 8.17
Breathing retraining should be limited to individuals who clearly hyperventilate during a
panic attack. In most cases breathing retraining will not be necessary. If it is included in
the treatment protocol, careful monitoring is needed to ensure it does not become a safetyseeking response.
324
TREATMENT OF SPECIFIC ANXIETY DISORDERS
TABLE 8.9. Diaphragmatic Breathing Retraining Protocol for Cognitive Therapy of Panic
PHASE I. BASELINE PREPARATION
Rationale: Review physical sensations and cognitions of most recent panic attack. Obtain belief
ratings at varying levels of anxiety to show how same sensations can be interpreted differently at
different times.
Instructions: Ask client to describe the physical sensations and the catastrophic misinterpretation
associated with the panic attack; client rates belief in the misinterpretation now and when most
anxious.
PHASE II. HYPERVENTILATION INDUCTION
Rationale: To demonstrate the production of physical sensations similar to a panic attack through
overbreathing.
Instructions: Individuals are asked to overbreathe at rate of 30 breaths per minute for 2 minutes or
until it becomes too difficult to continue. They are instructed in use of paper bag to rebreathe
expired CO2 . They are also instructed to focus on the physical sensations produced by
hyperventilation.
PHASE III. REATTRIBUTION
Rationale: To introduce possibility that physical sensations during panic are wrongly attributed to a
health threat and instead could be due to overbreathing.
Instructions: Clients are asked to review the physical sensations during hyperventilation and
the sensations described during panic. Rate their degree of similarity and discuss how the
overbreathing symptoms might be worse in a naturalistic setting.
PHASE IV. EDUCATION AND TREATMENT RATIONALE
Rationale: Explain the physiology of hyperventilation and how it causes uncomfortable physical
sensations
Instructions: Discuss how hyperventilation can cause an abrupt drop in arterial pCO2 that causes
uncomfortable physical sensations. When these symptoms are misinterpreted as indicating a
life-threatening danger like a heart attack, going crazy, or suffocation, panic sets in. Learning
to counter overbreathing with a slower, moderate breathing rate will reduce the intensity of the
physical sensations and provide new evidence that the sensations are due to overbreathing and
not the catastrophic health threat.
PHASE V. DIAPHRAGMATIC BREATHING
Rationale: Learn a relaxation coping skill to counter hyperventilation and other breathing
irregularities that cause the production of physical sensations that are misinterpreted in a
threatening manner.
Instructions:
1. Place one hand on chest with thumb just below neck and the other hand on stomach with little
finger just above naval.
2. Have client take short, shallow breaths through nose. Notice how hand on chest slightly rises
but hand on stomach hardly moves.
3. Now have client take slower, normal breaths through the nose and notice how the abdomen
moves slightly outward with each inhale and then deflates with each exhale.
4. Practice for 2–3 minutes with the client focusing on the movement of the abdomen with each
inhale and exhale.
5. Proceed to work on slowing the breathing rate down to 8 or 12 breaths per minute. Introduce a
paced breathing rate in which therapist demonstrates a 4 second inhale–4 second exhale cycle.
This can be done by counting 1–2–3–4 with each inhale, and then 1–2–3–4 with each exhale. A
short pause occurs at the end of each inhale and exhale. As the client exhales the word “relax”
should be repeated. After the therapist and client practice this slow, moderate breathing, the
client should continue with the diaphragmatic breathing with a particular focus on the slow
steady rhythm of breathing and the rise and fall of the stomach with each inhale and exhale.
6. Homework assignments begin with two to three daily 10-minute diaphragmatic breathing
practice sessions with or without a pacing audiotape. This is followed by daily sessions of 1–2
minutes of hyperventilation followed by slow breathing.
7. The final phase of homework involves application of diaphragmatic breathing in a variety of
anxious everyday situations.
Panic Disorder
325
EFFICACY OF COGNITIVE THERAPY FOR PANIC DISORDER
Cognitive behavior therapy for panic disorder falls within the American Psychological
Association’s well-established category of empirically supported treatments (Chambless et al., 1998; Chambless & Ollendick, 2001). The American Psychiatric Association Practice Guidelines for the treatment of panic disorder concluded that CBT was a
proven effective treatment for panic, with a 78% completer response rate that was at
least equal or superior to the effectiveness of antipanic medication (American Psychiatric Association, 1998).
Numerous reviews of the clinical outcome research have concluded there is strong
support for the efficacy of CBT for panic disorder. After reviewing more than 25 independently conducted clinical trials, Barlow and colleagues concluded that 40–90% of
patients treated with CBT are panic-free at end of treatment (Landon & Barlow, 2004;
White & Barlow, 2002). Other reviewers have also concluded that the effectiveness
of CBT for panic is strongly supported by the outcome literature (Butler, Chapman,
Forman, & Beck, 2006; DeRubeis & Crits- Christoph, 1998; Otto, Pollack, & Maki,
2000) and that treatment gains endure beyond termination more than with medication
(Hollon, Stuart, & Strunk, 2006). In the following section we provide a brief review of
selected key clinical outcome studies for CBT as well as dismantling studies that investigate the mechanism of change in the treatment package.
CBT Outcome Studies
Several meta-analyses have determined that CBT for panic is associated with superior
effect sizes. For example in a meta-analysis based on 13 studies Chambless and Peterman
(2004) obtained an average effect size of .93 for panic and phobic symptoms, with 71%
of CBT patients panic-free at posttreatment compared to 29% for the control conditions
(i.e., wait list or attention placebo). Furthermore, significant treatment gains were evident
in other symptom domains such as the cognitive symptoms of panic, generalized anxiety,
and, to a lesser extent, depression (see also Gould et al., 1995, for similar conclusions).
One of the earliest reports on cognitive therapy for panic disorder was a naturalistic
outcome study of 17 patients treated with a mean of 18 individual sessions of cognitive therapy that focused on misinterpretations of the physical symptoms of anxiety,
exposure, and cognitive restructuring of panic-relevant fears (Sokol, Beck, Greenberg,
Wright, & Berchick, 1989). At posttreatment panic frequency declined to zero and was
maintained at 1-year follow-up, and significant reductions were achieved on the BAI
and BDI. In addition improvement was made in patients’ ability to reappraise their fears
in a more realistic manner. In a later randomized clinical trial in which 33 patients with
panic disorder were assigned to 12 weeks of individual cognitive therapy or 8 weeks
of brief supportive psychotherapy, Beck, Sokol, Clark, Berchick, and Wright (1992)
found that at 8 weeks the cognitive therapy group had significantly fewer self-reported
and clinician-rated panic attacks than the comparison group. In addition the cognitive
therapy group had less generalized anxiety and fear but not less depression, and 71%
were panic-free compared to 25% in the psychotherapy condition. At 1-year follow-up
87% of the cognitive therapy group remained panic-free.
In a major outcome study 64 panic patients were randomly assigned to an average of
10 weekly individual sessions of cognitive therapy, applied relaxation, imipramine only,
326
TREATMENT OF SPECIFIC ANXIETY DISORDERS
or a 3-month wait list control followed by random assignment to one of the active treatments (D. M. Clark et al., 1994). At posttreatment (i.e., 3 months), cognitive therapy
was significantly more effective than applied relaxation and imipramine in reduction of
panic symptoms (i.e., panic composite score), agoraphobic avoidance, misinterpretation
of bodily sensations, and hypervigiliance for body symptoms. In addition 80% of the
cognitive therapy patients reached high-end functioning at 3 months compared to 25%
for applied relaxation and 40% for imipramine. Moreover, at 15-month follow-up cognitive therapy remained superior to applied relaxation and imipramine on six measures
of panic/anxiety, with 85% of cognitive therapy patients still panic-free compared with
47% of applied relaxation and 60% of imipramine patients.
In a large multisite randomized placebo-controlled clinical trial involving 77 patients
with panic disorder (Barlow, Gorman, Shear, & Woods, 2000) intent-to-treat analyses
revealed that CBT and imipramine were superior to placebo, but there were no significant differences between imipramine and CBT at posttreatment, although there was
a trend favoring CBT at 6-month follow-up. Overall, then, major treatment outcome
studies have clearly established that CBT for panic disorder is at least as effective as
medication, although there is little advantage in combining CBT with pharmacotherapy.
Comparisons of CBT with applied relaxation (i.e., Öst & Westling, 1995) indicate that
CBT is probably more effective for panic disorder (Siev & Chambless, 2007).
Outcome studies indicate that CBT can be effective for more difficult cases of panic
disorder. CBT can produce enduring treatment effects even with comorbid diagnoses, with significant improvement evident in both panic and comorbid symptoms (e.g.,
Craske et al., 2007; Tsao, Mystkowski, Zucker, & Craske, 2005). In fact Craske and
colleagues found more generalized symptom improvement in panic-focused CBT than
in a condition in which therapists were allowed to stray onto issues related to the comorbid condition. CBT has also been shown to be effective in drug-refractory individuals
with panic disorder (Heldt et al., 2006) and in reducing both day and nighttime panic
symptoms in patients with nocturnal panic attacks (Craske et al., 2005). Finally, brief
versions of CBT (e.g., intensive 2-day intervention), as well as computerized adaptations, can be highly effective for panic disorder (D. M. Clark et al., 1999; Deacon &
Abramowitz, 2006b; Kenardy et al., 2003). Although these findings are preliminary,
they do suggest that more efficient and cost-effective cognitive interventions may be
available for panic disorder.
CBT Process Studies
Exposure is an important component of cognitive therapy for panic disorder, especially
when agoraphobic avoidance is prominent. Given our emphasis on cognitive intervention, how critical is cognitive restructuring to the effectiveness of CBT for panic disorder? In their meta-analysis, Gould et al. (1995) found that cognitive restructuring plus
interoceptive exposure (i.e., symptom induction or schema activation) yielded the largest
effect sizes, but cognitive restructuring alone produced highly variable results. In an
early study Margraf and Schneider (1991) found cognitive restructuring without exposure as effective as pure exposure or combined exposure plus cognitive restructuring.
In a series of multiple baseline single cases Salkovskis et al. (1991) found that two
sessions of cognitive restructuring focused on evidence gathering for and against the
Panic Disorder
327
patient’s catastrophic interpretation of physical symptoms produced significant reduction in panic frequency in six out of seven patients, whereas nonfocal treatment had
little effect on panic symptoms. In a more recent multivariate time series single-case
analysis both cognitive restructuring with empirical hypothesis testing versus exposure alone produced equivalent changes in dysfunctional beliefs and self-efficacy that
preceded improvements in panic apprehension (Bouchard et al., 2007). The authors
concluded that the findings add to growing empirical evidence that cognitive changes
precede improvement in panic symptoms whether treatment is primarily cognitive or
behavioral. Other studies have found that exposure alone is as effective as exposure
plus cognitive restructuring (Bouchard et al., 1996; Öst, Thulin, & Ramnerö, 2004),
although Van den Hout, Arntz, and Hoekstra (1994) found that cognitive therapy alone
reduced panic attacks but not agoraphobia. In a recent study of group CBT for panic,
20% of patients experienced a sudden gain (i.e., rapid symptom reduction) after two
sessions and this predicted better symptom outcome at posttreatment (Clerkin, Teachman, & Smith-Janik, 2008). Overall, these studies indicate that CBT can produce rapid
and effective symptom reduction in panic disorder and that cognitive restructuring is an
important component of the treatment package.
The therapeutic effects of cognitive restructuring suggest that targeting the catastrophic misinterpretations of bodily sensations is a central mechanism of change in
cognitive therapy of panic disorder. In their clinical trial D. M. Clark at al. (1994) found
a significant correlation between BSIQ scores at 6 months and panic symptoms and
relapse rates at 15 months. This relation between a continued tendency to misinterpret
bodily sensations and worst outcome at follow-up was supported in the authors’ outcome study of brief cognitive therapy (D. M. Clark et al., 1999). However, comparison
of standard cognitive therapy that focused on interpersonal beliefs relevant to generalized anxiety versus focused cognitive therapy that targeted catastrophic misinterpretations of bodily sensations showed that both were equally effective in reducing panic
symptoms, although reduction in panic-related cognitions and beliefs was correlated
with changes in panic frequency at termination (Brown, Beck, Newman, Beck, & Tran,
1997). In their descriptive and meta-analytic review of 35 CBT studies on panic disorder, Oei, Llamas, and Devilly (1999) concluded that the therapy is effective for panic
disorder and does produce change in cognitive processes, although it is unclear whether
change in catastrophic misinterpretations is the central change mechanism in CBT for
panic disorder. Overall it would appear that change in catastrophic misinterpretations
of the physical symptoms of anxiety is an important part of the treatment process in
panic but whether a specific focus on these symptoms is necessary remains unclear.
Clinician Guideline 8.18
Cognitive therapy involving cognitive restructuring, symptom induction, and empirical
hypothesis-testing exposure exercises is a well-established empirically based treatment for
panic disorder with or without agoraphobic avoidance. Cognitive strategies and exposureoriented homework are both central ingredients in the treatment’s efficacy for panic
attacks.
328
TREATMENT OF SPECIFIC ANXIETY DISORDERS
SUMMARY AND CONCLUSION
The problem of recurrent panic attacks provides the clearest example of the cognitive
conceptualization of fear. Occurrence of at least two unexpected panic attacks, apprehension or worry about further attacks, and avoidance of situations thought to trigger
panic are hallmarks of panic disorder.
A revised cognitive model of panic disorder was presented in Figure 8.1. The essential components of this model are (1) increased attention or hypervigilance for certain
physical or mental sensations, (2) activation of physiological or mental threat schemas,
(3) the catastrophic misinterpretation of physical symptoms as indicating an imminent
dire threat to self, (4) further intensification of the physical symptoms of anxiety, (5) loss
of ability to reappraise symptoms in a more realistic, benign manner, and (6) reliance
on avoidance and safety seeking to reduce heightened anxiety and terminate the panic
episode. Empirical evidence, reviewed for the model’s six key hypotheses, found strong
support for increased responsiveness to internal states, the activation of prepotent physiological or mental threat schemas, the catastrophic misinterpretation of bodily sensations, and the functional role of avoidance and safety seeking in the persistence of panic
attacks.
Table 8.7 summarized the main components of cognitive therapy for panic disorder.
Reduction in hypervigilance for feared bodily sensations, reversal of the catastrophic
misinterpretation of internal states, increased ability to produce more realistic and balanced reappraisals of the feared symptoms of anxiety, reduction in avoidance and safety
seeking, and an increased sense of safety are the primary goals of cognitive therapy.
These are achieved using within-session symptom induction to activate threat schemas,
cognitive restructuring to weaken catastrophic misinterpretations and improve reappraisal capacity, and systematic situational and interoceptive exposure assignments in
a hypothesis-testing context. Over the last two decades a number of well-designed randomized clinical trials have established cognitive therapy as a highly efficacious treatment for panic disorder with or without agoraphobic avoidance.
There are a number of issues that remain for cognitive theory and therapy of panic
disorder. Panic disorder is characterized by increased responsiveness to changes in internal state, although the specific processes that contribute to this interoceptive hypersensitivity are not well understood. It is still not clear whether a catastrophic misinterpretation of bodily sensations is necessary for the production of all panic attacks, whether it
is a cause or a consequence of repeated panic attacks, and whether the concept should
be broadened to include imminent social and emotional threats such as fear of further
panic attacks. Moreover, there is insufficient research on whether loss of reappraisal
capacity is a major determinant of recurrent panic attacks and the role played by panic
self-efficacy or perceived effectiveness in terminating panic episodes. In terms of treatment effectiveness, comparative outcome studies of cognitive therapy versus the newer
SSRIs are needed as well as longer follow-up periods to determine the enduring benefits
of treatment. Nevertheless, cognitive therapy/CBT is now considered a well- established
and efficacious treatment for panic disorder with or without agoraphobia and should be
the first-line treatment choice in most cases of the disorder.
APPENDIX 8.1
Weekly Panic and Acute Anxiety Log
Name:
Date:
Instructions: Please use this form to record any panic attacks, limited panic attacks or acute anxiety episodes that you experienced in the past
week. Try to complete the form as close to the anxiety episode as possible in order to increase the accuracy of your remarks.
Severity/Intensity
Date, Time,
of Anxiety (0–100); Description of the
and Duration Situational Triggers (Label FPA, LPA, Anxious Physical or
of Episode
(Label E or UE)*
AAE)*
Mental Sensations
Anxious
Interpretation of
the Sensations
Evidence for
an Alternative
Interpretation of
Sensations
Outcome (Coping
Responses and
Sense of Safety)+
1.
329
2.
3.
4.
5.
* E = expected to have panic in this situation; UE = panic occurred unexpectedly, completely out of the blue; FPA = full-blown panic attack; LPA = limited symptom attack;
AAE = acute anxiety episode (sudden onset of anxiety but not panic)
+ rate sense of safety after panic ceases from 0 = don’t feel at all safe from panic to 100 = feel absolutely safe from further panic
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright 2010 by The Guilford Press. Permission to photocopy
this appendix is granted to purchasers of this book for personal use only (see copyright page for details).
APPENDIX 8.2
Symptom Reappraisal Form
Name:
Date:
Instructions: Please use this form to write down any alternative explanations you can think for why you are experiencing a variety of physical
sensations that make you feel anxious or panicky.
State the Physical
Sensation Experienced
(e.g., racing heart,
breathlessness, nausea)
List a Number of Alternative Explanations for the
Sensations Other Than the Worst Outcome (i.e., the
feared catastrophe)
Rate Belief in
Rate Belief in
Effectiveness
Each Explanation Each Explanation
of Explanation
When Not
When in Anxious
in Countering
Anxious (0–100)* State (0–100)* Anxiety (0–100)+
1.
330
2.
3.
4.
5.
* For belief ratings, 0 = absolutely no belief in the explanation, 100 = absolutely certain that this is the cause of the physical sensations.
+ For effectiveness ratings, 0 = explanation has absolutely no positive effect on anxiety, 100 = explanation is completely effective in eliminating anxious feelings.
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright 2010 by The Guilford Press. Permission to photocopy
this appendix is granted to purchasers of this book for personal use only (see copyright page for details).
APPENDIX 8.3
Vicious Cycle of Panic
Name:
Date:
Situational Triggers
1.
2.
3.

Initial Physical, Mental, Emotional Symptoms
1.
2.
3.

First Anxious (Apprehensive) Thoughts/Images
1.
2.
3.

Main Escalating Symptoms
1.
2.

Thoughts/Images of Imminent Danger (catastrophe)
Attempts to Cope/Control

PANIC
From Cognitive Therapy of Anxiety Disorders: Science and Practice by David A. Clark and Aaron T. Beck. Copyright
2010 by The Guilford Press. Permission to photocopy this appendix is granted to purchasers of this book for personal
use only (see copyright page for details).
331
Chapter 9
Cognitive Therapy of Social Phobia
Nothing so much prevents our being natural
as the desire to seem so.
—F RANÇOIS , DUC DE L A ROCHEFOUCAULD
(French writer and aristocrat, 1613–1680)
Gerald is a 36-year-old man who has worked as an accountant for a large
multinational trucking firm for the past 12 years and who has a long history of
severe social anxiety. SCID assessment revealed that he met DSM-IV diagnostic criteria for generalized social phobia. He reported intense anxiety in most
social situations with an overwhelming fear that other people will notice him.
His main concern was that they would notice that he was hot and flushed and
would think “What’s wrong with him?”, “He doesn’t look normal,” and “He
must have low self-esteem or some serious mental problem.” Gerald believed
that people “could look right through him” and so he was always hypervigilant
when around others. He was also concerned that others would think he was
boring and wasting their time. Gerald noted that he is almost always anxious
when around other people and recognized that his anxiety is excessive. Over
the years he got to the point where he avoids social contact as much as possible,
spending most of his time outside of work alone and isolated. He has never
had an intimate relationship and no close friends. He prefers to avoid people
because of the anxiety and a fear that social interaction will result in obligations to others even though he realizes the avoidance has been detrimental to
his career. A year ago he joined a health club for a few months but found it too
anxiety-provoking to attend. Gerald rated participating in meetings, taking a
course, meeting an unfamiliar person, answering the telephone, taking public
transportation, or even visiting an acquaintance as very anxiety-provoking.
Gerald indicated that he has been socially anxious since childhood and that
it has severely limited his life. In fact the anxiety and self-imposed loneliness
have been so great that he commented “I’m tired of waiting for life to start;
sometimes I just want to get it over with.”
Gerald’s clinical presentation is fairly typical of someone with a chronic
and severe generalized social phobia. In fact he met criteria for an Axis II
avoidant personality disorder as indicated by (1) his attempt to avoid signifi332
Social Phobia
333
cant interpersonal contact at work (he would start work at 7:00 A.M. and quit
at 2:00 P.M. in order to minimize contact with others), (2) unwillingness to
get involved with people, (3) fear of intimate relationships, (4) inhibition in
new interpersonal relationships because of inadequacy feelings, (5) perceived
inferiority to others, and (6) reluctance to engage in any new, even relatively
mundane, social activities for fear of embarrassment. Gerald received 19 sessions of cognitive therapy that focused specifically on his social evaluative anxiety, inhibitory behavior, and extreme avoidance. Therapy targeted Gerald’s
maladaptive beliefs about negative social evaluation by others, his reliance on
escape and avoidance to manage anxiety, and graded in vivo exposure to moderately anxious social situations.
This chapter presents the cognitive theory and treatment of generalized social phobia as first described in Beck et al. (1985, 2005). We begin with a discussion of the
diagnostic and phenomenological characteristics of social phobia. This is followed by
a description of a more elaborated cognitive theory of social phobia as well as a review
of its empirical support. We then propose a cognitive approach to assessment and treatment of social phobia. The chapter concludes with a review of the empirical status of
cognitive therapy and CBT for generalized social phobia.
DIAGNOSTIC CONSIDERATIONS
Diagnostic Overview
The core feature of social phobia is a “marked and persistent fear of social or performance situations in which embarrassment may occur” (DSM-IV-TR; APA, 2000,
p. 450). Although anxious feelings are common to most people when they enter novel,
unfamiliar, or social-evaluative situations like a job interview, the person with social
phobia invariably experiences intense fear or dread, even when anticipating the possibility of exposure to various common social situations. The anxiety stems from a fear of
scrutiny and negative evaluation from others that will lead to feelings of embarrassment,
humiliation, and shame (Beck et al., 1985, 2005). The perceived cause of the embarrassment usually centers on some aspect of self-presentation such as exhibiting a symptom(s)
of anxiety, speaking awkwardly, making a mistake, or acting in some other humiliating
manner (Heckelman & Schneier, 1995). As a result the person with social phobia tends
to be highly self- conscious and self- critical in the feared social situation, often exhibiting involuntary inhibitory behaviors such as appearing stiff and rigid or being verbally
inarticulate, which results in detrimental social performance and the unwanted attention of others.
Social phobia is closely related to simple phobia because the fear occurs only in
situations in which the person must do something in the context of being observed and
possibly evaluated by others (Hofmann & Barlow, 2002). The person with social phobia
who experiences intense anxiety while eating, speaking, or writing in front of unfamiliar people has no difficulty engaging in these behaviors when alone or with family and
close friends. Although Marks and Gelder (1966) first described the syndrome of social
phobia (see also Marks, 1970), it was not until DSM-III (American Psychiatric Association, 1980) that it was incorporated as a separate diagnostic entity. The core diagnostic
criteria have remained constant throughout subsequent DSM revisions with the excep-
334
TREATMENT OF SPECIFIC ANXIETY DISORDERS
tion that a generalized subtype of social phobia was introduced in DSM-III-R (American Psychiatric Association, 1987) and the exclusionary rule for avoidant personality
disorder was removed. Although an alternate label, social anxiety disorder, has been
recommended (Liebowitz, Heimberg, Fresco, Travers, & Stein, 2000), we retain use of
the term “social phobia” because it captures the strong urge to avoid anxiety-provoking
situations that is the hallmark of the disorder. Table 9.1 presents the DSM-IV-TR diagnostic criteria for social phobia.
Fear of Negative Evaluation
Fear of negative evaluation by others is a core feature of social phobia that is not only
recognized in cognitive models of the disorder (Beck et al., 1985, 2005; D. M. Clark,
2001; Rapee & Heimberg, 1997; Wells & Clark, 1997), but it is the basis of the marked
and persistent fear in social evaluative situations described in DSM-IV-TR Criterion
A. Individuals with social phobia may hold excessively high standards of social performance, wanting to make a particular impression on others but doubting their ability
to actually make a positive impression (Beck et al., 1985, 2005; Hofmann & Barlow,
2002). They also believe they draw the attention of others in social situations and live
in fear that in this social evaluative context they will embarrass or humiliate themselves
TABLE 9.1. DSM-IV-TR Diagnostic Criteria for Social Phobia
A. A marked and persistent fear of one or more social or performance situations in which the
person is exposed to unfamiliar people or to possible scrutiny by others. The individual
fears that he or she will act in a way (or show anxiety symptoms) that will be humiliating or
embarrassing.
B. Exposure to the feared social situation almost invariably provokes anxiety, which may take the
form of a situationally bound or situationally predisposed Panic Attack.
C. The person recognizes that the fear is excessive or unreasonable.
D. The feared social or performance situations are avoided or else are endured with intense anxiety
or distress.
E. The avoidance, anxious anticipation, or distress in the feared social or performance situation(s)
interferes significantly with the person’s normal routine, occupational (academic) functioning, or
social activities or relationships, or there is marked distress about having the phobia.
F. In individuals under 18 years of age, the duration is at least 6 months.
G. The fear or avoidance is not due to the direct physiological effects of a substance (e.g., drug
of abuse, medication), or a general medical condition, and is not better accounted for by
another mental disorder (e.g., Panic Disorder With or Without Agoraphobia, Separation
Anxiety Disorder, Body Dysmorphic Disorder, a Pervasive Developmental Disorder, or Schizoid
Personality Disorder).
H. If a medical condition or another mental disorder is present, the fear in Criterion A is unrelated
to it (e. g., the fear is not of Stuttering, trembling in Parkinson’s disease, or exhibiting abnormal
eating behavior in Anorexia Nervosa or Bulimia Nervosa).
Specify if:
Generalized: if the fears include most social situations (also consider the additional diagnosis of
Avoidant Personality Disorder)
Note. From American Psychiatric Association (2000). Copyright 2000 by the American Psychiatric Association.
Reprinted by permission.
Social Phobia
335
by acting or appearing foolish, less intelligent, or visibly anxious (Beidel & Turner,
2007). There is considerable empirical evidence that fear of negative evaluation is a core
feature of social phobia (e.g., Ball et al., 1995; Hackmann et al., 1998; Hirsch & Clark,
2004; Mansell & Clark, 1999; Voncken, et al., 2003). However, individuals with social
phobia may fear any social evaluation, either positive or negative, that involves feelings
of conspicuousness or self- consciousness (Weeks, Heimberg, Rodebaugh, & Norton,
2008). Moreover, the negative evaluation feared by those with social phobia is not simply some mildly negative impression on others but a much more extreme experience of
dreaded humiliation or shame (Beck et al., 1985, 2005). Shame is a painful affect in
which personal attributes, characteristics, or behavior are perceived as causing a loss
of social standing or attractiveness to others, or even worse, their outright criticism or
rejection (Gilbert, 2000).
Social Situations
The majority of individuals with social phobia experience marked anxiety in a variety
of social situations (Rapee, Sanderson, & Barlow, 1988; Turner, Beidel, Dancu, & Keys,
1986). Rachman (2004) noted that the most common situations feared in social phobia are public speaking, attendance at parties or meetings, and speaking to authority
figures. Beidel and Turner (2007) reported that formal speaking (the most distressing
situation), parties, initiating and maintaining conversations, and informal speaking and
meetings were rated as distressing and avoided by more than 75% of patients with social
phobia. Dating was rated as distressing and something avoided by half of the sample,
whereas eating and drinking in public, using public washrooms, and writing in public
was feared by 25% or less of social phobic individuals. The anxiety-provoking situations in social phobia have been categorized as those dealing with social interaction
versus those concerned with performance (Rapee, 1995b). Table 9.2 presents a list of
interpersonal and performance situations from Antony and Swinson (2000b) that are
rated for level of fear and avoidance when assessing social phobia.
TABLE 9.2. Common Interpersonal and Performance Situations Feared in Social Phobia
Interpersonal situations
Performance situations
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
Initiating a date or appointment with someone
Being introduced to unfamiliar person
Attending a party or social gathering
Having friend for dinner
Starting a conversation
Talking on phone to a familiar person
Talking on phone to an unfamiliar person
Expressing your personal opinion to others
Having job interview
Being assertive with others
Returning a purchased item
Making eye contact
Expressing dissatisfaction with restaurant food
Talking to authority figures
Note. Based on Antony and Swinson (2000b).
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
•฀
Making a toast or speech
Speaking in meetings
Playing sports in front of an audience
Participating in a wedding party or public
ceremony
Singing/performing to an audience
Eating/drinking in a public setting
Using public washrooms
Writing in front of others
Making a mistake in public (e.g.,
mispronouncing a word)
Walking/running in busy public place
Introducing yourself to others
Shopping in a busy store
Walking in front of a large group of people (e.g.,
walking up aisle of church, theater)
336
TREATMENT OF SPECIFIC ANXIETY DISORDERS
Anxious Arousal and Panic
The second diagnostic criterion in DSM-IV-TR is that exposure to the feared social situation will invariably provoke anxiety, which may involve a situationally bound or situationally predisposed panic attack (American Psychiatric Association, 2000). Individuals with social phobia often experience panic attacks when in feared social situations or
even when anticipating a social event (Kendler, Neale, Kessler, Heath, & Eaves, 1992c).
Although the physical symptoms of these situationally triggered attacks are identical to
those in panic disorder (Beidel & Turner, 2007), physical symptoms of anxiety that can
be observed by others, such as twitching muscles or blushing, may be more prominent in
the anxiety experienced in social phobia (Amies, Gelder, & Shaw, 1983). Furthermore,
individuals with social phobia experience greater physiological arousal during exposure
to a distressing social situation than nonphobic individuals (e.g., Turner et al., 1986). It
is little wonder that fear of having a panic attack in a social situation is a major concern
of many people with social phobia (Hofmann, Ehlers, & Roth, 1995). In fact fear of
losing control over any emotional responses, especially anxiety symptoms, is a critical
aspect of the perceived social threat (Hofmann, 2005). Even though fear of anxiety
is common across the anxiety disorders, it is particularly germane to social phobia
because any display of anxiety in social settings is perceived to increase the likelihood
of negative evaluation by others.
Awareness, Avoidance, and Inhibition
To meet DSM-IV-TR diagnostic criteria for social phobia, the person must have some
awareness of the excessive or unreasonable nature of her social fears (i.e., Criterion C).
This criterion helps distinguish social phobia from other diagnoses such as paranoid
personality disorder in which the person actually believes others are trying to embarrass
or humiliate him (Beidel & Turner, 2007).
Given the experience of intense anxiety when anticipating or entering feared social
situations, the urge to avoid social situations can be intense in social phobia. Compared
to other anxiety disorders, individuals with social phobia are more likely to engage in
avoidance of social situations even though they may be convinced it is detrimental for
them (i.e., Rapee, Sanderson, & Barlow, 1988). Assessment of the frequency and extent
of avoidance associated with various social evaluative situations (see Table 9.2) is an
important part of the diagnostic assessment of social phobia (Hope, Laguna, Heimberg,
& Barlow, 1996–1997).
Individuals with social phobia are highly inhibited when encountering social interactions. They often appear rigid and stiff, their face taut with forced expression. When
trying to speak they can appear inarticulate because of stumbling over their words,
being “tongue-tied,” or having difficulty thinking of the right word. All of these involuntary behaviors are detrimental to their performance and increase the probability of a
negative evaluation by others—the very essence of their social anxiety.
Individuals with social phobia also rely on subtle avoidance or safety behaviors in
an attempt to conceal their anxiety which they assume will cause others to evaluate them
negatively (Beck et al., 1985; Wells & Clark, 1997). Individuals with social phobia may
try to conceal their anxiety by avoiding eye contact or trying to keep physically cool so
that one’s face does not look red or flushed, wearing certain clothes or makeup to hide
Social Phobia
337
blushing, give an excuse for one’s red face by blaming it on a hot room or not feeling well,
or the like (D. M. Clark, 2001). These concealment strategies (i.e., safety behaviors) are
problematic because they can directly exacerbate anxious symptoms (e.g., person wears
a heavy sweater to conceal sweating but this raises body temperature and tendency to
sweat). In addition the behaviors prevent disconfirmation of the feared outcome (e.g.,
attributes nonoccurrence of negative evaluation to performance of the safety behavior),
maintain a heightened self-focused attention, and draw greater negative attention from
others (Wells & Clark, 1997). There is some evidence that socially anxious individuals realize the negative social effects of trying to conceal anxiety (Voncken, Alden, &
Bögels, 2006) but still tend to engage in safety behaviors (Alden & Bieling,1998).
Marked Distress and Interference
Anxiety or nervousness in social situations is common in the general population. In a
randomized community telephone survey of 526 adults, 61% reported feeling nervous
or uncomfortable in at least one of seven social situations with public speaking being
the most frequently endorsed situation (Stein, Walker, & Forde, 1994). Thus the marked
distress or interference criterion in DSM-IV-TR is needed to distinguish the more severe,
clinical forms of social phobia disorders from the milder subclinical variants of social
anxiety found throughout the nonclinical population (Heckelman & Schneier, 1995).
Clinician Guideline 9.1
Social phobia is characterized by a marked and persistent anxiety, even panic, most often
across numerous interpersonal and/or performance situations in which the person fears
scrutiny and negative evaluation by others that will lead to embarrassment, humiliation,
or shame. A key concern is that one’s interpersonal behavior, appearance, or expression of
anxiety will be negatively judged by others. Anticipatory anxiety can be intense, leading to
extensive avoidance of feared social situations, as well as production of involuntary inhibitory responses and attempts to conceal anxiety when social interaction is unavoidable.
Shyness and Social Phobia
There is considerable confusion about the relationship between shyness and social phobia, with some emphasizing their common characteristics of high social anxiety and
fear of negative evaluation by others (Stravynski, 2007), whereas others note there are
important quantitative differences so that the two should not be considered synonymous
(Bruch & Cheek, 1995). Like social phobia, shyness has been described as anxiety, discomfort, and inhibition in social situations and fear of negative evaluation by others,
especially authority figures (Heiser, Turner, & Beidel, 2003). Some have concluded that
social phobia is very similar to chronic shyness (Henderson & Zimbardo, 2001; Marshall & Lipsett, 1994). Moreover, delineating clear boundaries between shyness and
social phobia has been difficult because (1) there is no consensus on the definition of
shyness; (2) they have many shared behavioral, cognitive, and physiological features; (3)
they arise from different research traditions with shyness studied by social, personality,
338
TREATMENT OF SPECIFIC ANXIETY DISORDERS
and counseling psychologists whereas social phobia is a research topic in clinical psychology; and (4) their differences may be more quantitative than qualitative in nature
(Bruch & Cheek, 1995; Heckelman & Schneier, 1995; Rapee, 1995b).
Shyness is a normal personality trait that involves some degree of nervousness,
inhibition, and self- consciousness in social interaction. Butler (2007) described shyness
as a sense of shrinking back from social encounters and retreating into one’s self due
to physical discomfort (e.g., tension, sweating, trembling), feeling anxious, inhibition,
or inability to express yourself, and excessive self-focused attention. Zimbardo defined
shyness as “a heightened state of individuation characterized by excessive egocentric
preoccupation and overconcern with social evaluation . . . with the consequence that the
shy person inhibits, withdraws, avoids, and escapes” (cited in Henderson & Zimbardo,
2001, p. 48). Despite the similarities with social phobia, there are important differences.
Compared to social phobia, shyness is much more pervasive in the general population,
it may be less chronic or enduring, it is associated with less avoidance and functional
impairment, and shy individuals may be more able to engage in social interaction when
necessary (Beidel & Turner, 2007; Bruch & Cheek, 1995).
Table 9.3 presents some important differences between shyness and social phobia
(Turner, Beidel, & Townsley, 1990).
Social phobia is undoubtedly a more severe condition than shyness, with severe
and pervasive avoidance of social situations being one of the most important distinctions. Although the differences are more quantitative than qualitative in nature (Rapee,
1995b), Beidel and Turner (2007) concluded in their review that social phobia should
not be considered an extreme form of shyness. Studies that have directly compared
prevalence of shyness and social phobia confirm the distinctiveness of the two syndromes. Chavira, Stein, and Malcane (2002) found that only 36% of individuals who
had high levels of shyness met criteria for generalized social phobia compared to 4%
of individuals with average or normative shyness. In another study only 17.7% of shy
university students met diagnostic criteria for social phobia (Heiser et al., 2003) and
analysis of the NCS data set revealed a lifetime prevalence for social phobia of 28% for
women and 21% for men who reported excessive childhood shyness (Cox, MacPherson,
& Enns, 2005). Conversely only 51% of women and 41% of men with lifetime complex
(generalized) social phobia had excessive childhood shyness. Together these findings
TABLE 9.3. Distinguishing Features of Shyness and Social Phobia
Shyness
Social phobia
•฀ Normal personality trait
•฀ Primarily social inhibition and reticence
•฀ Psychiatric disorder
•฀ Presence of marked anxiety, even panic, in social
evaluative situations
•฀ More likely to exhibit poor social performance
•฀ Avoidance of social situations more frequent and
pervasive
•฀ Lower prevalence rate
•฀ Longer duration, more chronic, and unremitting
•฀ Later onset in early to midadolescence
•฀ Greater social and occupational impairment
•฀ Can socially engage when necessary
•฀ Less likely to avoid social situations
•฀
•฀
•฀
•฀
Highly prevalent in population
More transitory course for many individuals
Earlier onset perhaps in preschool years
Less impairment in daily living
Social Phobia
339
indicate that shyness and social phobia, though significantly related, can not be considered synonymous.
Social Phobia Subtypes: Generalized versus Specific
DSM-IV-TR (APA, 2000) allows for the distinction between a generalized and a specific
or circumscribed subtype of social phobia. Unfortunately the criteria for making this
distinction are not at all clear. Generalized social phobia (GSP) can be specified when
individuals fear most social situations including both public performance and social
interaction situations. However, the number of feared situations needed to qualify for
GSP is not stated. The “specific subtype” of social phobia is even less clearly defined.
DSM-IV-TR states that this subtype may be quite heterogeneous including people who
fear just a single performance situation (i.e., public speaking) as well as those who fear
most performance situations but not social interaction situations. In their prospective
community study, Wittchen, Stein, and Kessler (1999) reported a lifetime prevalence of
5.1% for specific social phobia and 2.2% for GSP among 14- to 24-year-old, with the
specific subtype mostly characterized by fears of test performance and speaking in front
of people.
There is considerable debate in the literature on the validity of the generalized versus specific distinction in social phobia. Not only are the DSM-IV-TR descriptions of
generalized and specific social phobia ambiguous, but researchers employ different definitions of the specific subtype from reserving the term for fear of public speaking only
to a broader definition that includes fear of multiple situations within one social domain
such as social performance situations only (see Hofmann & Barlow, 2002). Furthermore, a more fundamental problem for subtyping is that social phobia appears to lie on
a continuum of severity with no clear-cut boundaries to delineate subtypes. Taxometric
analyses indicate that social anxiety favors a dimensional model of severity (Kollman,
Brown, Liverant, & Hofmann, 2006) and community-based studies have failed to find
a clear demarcation of subtypes based on the number of feared social situations (e.g.,
Stein, Torgrud, & Walker, 2000; Vriends, Becker, Meyer, Michael, & Margraf, 2007a).
These findings suggest that the generalized distinction may be confounded with symptom severity so that the specifier may be arbitrarily selecting out the most severe on the
social anxiety continuum.
Others, however, have argued that specifying a generalized subtype is a clinically
useful distinction. The majority of individuals with social phobia who seek treatment
will meet criteria for the generalized subtype (see Beidel & Turner, 2007; e.g., Kollman
et al., 2006), whereas specific social phobia may be more prevalent in community samples (Wittchen et al., 1999). In addition GSP is associated with greater symptom severity, depression, avoidance, and fear of negative evaluation, as well as greater functional
impairment, earlier onset, greater chronicity, and increased rate of comorbid Axis I and
II diagnoses (e.g., Herbert, Hope, & Bellack, 1992; Holt, Heimberg, & Hope, 1992;
Kessler, Stein, & Berglund, 1998; Mannuzza et al., 1995; Turner, Beidel, & Townsley,
1992; Wittchen et al., 1999). Overall, the findings indicate that the generalized versus
specific subtype of social phobia is really capturing a severity distinction based on the
number of feared social situations, with GSP the more severe form of social phobia
that is most often seen in treatment settings. For this reason the cognitive perspective
described in this chapter is most relevant to GSP.
340
TREATMENT OF SPECIFIC ANXIETY DISORDERS
Clinician Guideline 9.2
Rather than forming distinct subtypes, social phobia varies along a continuum of severity
with milder forms involving fear of a limited range of social situations and more severe, generalized social phobia characterized by fear of a wider number of both social interaction and
performance situations.
Social Phobia and Avoidant Personality Disorder
A high degree of diagnostic overlap exists between GSP and avoidant personality disorder (APD) which has led researchers to question whether they really are two separate
conditions as currently described in DSM-IV (Sanderson, Wetzler, Beck, & Betz, 1994;
Tyrer, Gunderson, Lyons, & Tohen, 1997; van Velzen, Emmelkamp, & Scholing, 2000;
Widiger, 1992). As can be seen from the diagnostic criteria for APD in Table 9.4, both
GSP and APD share much in common because essentially both are characterized by
a pervasive pattern of discomfort, inhibition, and fear of negative evaluation across a
variety of social or interpersonal contexts (Heimberg, 1996).
In his review Heimberg (1996) concluded that approximately 60% of individuals
with GSP will meet criteria for APD compared with 20% of nongeneralized social phobias. Moreover, almost all individuals with APD will meet diagnostic criteria for social
phobia (Brown, Heimberg, & Juster, 1995; Herbert et al., 1992; Turner et al., 1992).
Given this close relationship between GSP and APD, an assessment for APD should be
made whenever individuals meet diagnostic criteria for social phobia.
Comparison of the clinical presentation between GSP with and without APD have
generally revealed that those with GSP and APD have greater symptom severity, diagnostic comobidity, functional impairment, social skills deficiencies, and possibly less
motivation and response to CBT (e.g., Holt et al., 1992; van Velzen et al., 2000; see
also Beidel & Turner, 2007; Heimberg, 1996). More recently Chambless, Fydrich, and
TABLE 9.4. DSM-IV-TR Diagnostic Criteria for Avoidant Personality Disorder
A pervasive pattern of social inhibition, feelings of inadequacy, and hypersensitivity to negative
evaluation, beginning by early adulthood and present in a variety of contexts, as indicated by four
(or more) of the following:
(1) avoids occupational activities that involve significant interpersonal contact, because of fears of
criticism, disapproval, or rejection
(2) is unwilling to get involved with people unless certain of being liked
(3) shows restraint within intimate relationships because of the fear of being shamed or ridiculed
(4) is preoccupied with being criticized or rejected in social situations
(5) is inhibited in new interpersonal situations because of feelings of inadequacy
(6) views self as socially inept, personally unappealing, or inferior to others
(7) is unusually reluctant to take personal risks or to engage in any new activities because they may
prove embarrassing
Note. From American Psychiatric Association (2000). Copyright 2000 by the American Psychiatric Association.
Reprinted by permission.
Social Phobia
341
Rodebaugh (2006) found that GSP with APD was characterized by a more severe form
of social phobia and poorer social skills compared to GSP without APD, with low selfesteem in the APD group the only qualitative difference. The authors concluded that
DSM-IV APD should be considered a severe form of GSP rather than a separate diagnosis.
Beidel and Turner (2007) raise a number of treatment implications that may argue
for the clinical utility of retaining the APD diagnosis. They note that individuals with
APD may have less tolerance for exposure-based treatment and so a more gradual
approach may be necessary. They also indicate that individuals with APD may have
more social skills deficits and lower social/occupational functioning, thereby making
social skills training an essential treatment ingredient when APD is present. Despite
these clinical observations, the empirical research to date suggests that social anxiety should be conceptualized as a continuum of severity with specific or circumscribed
social phobia at the milder end, GSP without APD in the moderate range, and GSP with
APD the most severe form of the disorder (McNeil, 2001).
Clinician Guideline 9.3
Avoidant personality disorder (APD) is a severe form of GSP associated with greater psychopathology and functional impairment. Given the treatment complications that may be
associated with this diagnosis, include an assessment of APD in the diagnostic protocol for
social phobia.
EPIDEMIOLOGY AND CLINICAL FEATURES
Prevalence
Social phobia is the most common of the anxiety disorders and third most common
over all mental disorders. The NCS employed DSM-III-R criteria for social phobia and
found that 12-month prevalence was 7.9% and lifetime prevalence was 13.3% (Kessler et al., 1994). Moreover, approximately two-thirds of these individuals had GSP,
with the remainder having purely speaking fears that were less persistent and impairing
(Kessler et al., 1998). The more recent NCS-R based on DSM-IV diagnostic criteria
reported a 12-month prevalence of 6.8% and lifetime prevalence of 12.1% for social
phobia (Kessler, Berglund, et al., 2005; Kessler, Chiu, et al., 2005). The high prevalence
for social phobia has been found in other epidemiological and large community studies (e.g., Newman et al., 1996). There is also some evidence that the incidence of social
phobia may be increasing over time (Rapee & Spence, 2004).
As noted previously, milder forms of social anxiety are more prevalent in the general
population than social phobia. Social inhibition, fear of negative evaluation, anxiousness, and feelings of inadequacy when in social situations are reported to occur occasionally to moderately often by the majority of nonclinical individuals. Moreover, fear
and avoidance of social situations is common in panic disorder, GAD, and agoraphobia
(Rapee et al., 1988). What distinguishes social phobia is the number of social situations
feared and the degree of functional impairment (Rapee et al., 1988; Stein et al., 2000).
342
TREATMENT OF SPECIFIC ANXIETY DISORDERS
Gender and Cross-Cultural Differences
Unlike the other anxiety disorders, the gender ratio for social phobia is not as highly
skewed toward women. There is an approximate 3:2 ratio of women to men with social
phobia. In the NCS the lifetime prevalence for women was 15% and for men 11.1%
(Kessler et al., 1994). However, Rapee (1995b) notes that an equal number of men and
women seek treatment for social phobia, although nonclinical questionnaire studies
suggest women may feel greater social anxiety and shyness than men (e.g., Wittchen et
al., 1999). Cross- cultural differences may also be apparent in the gender ratio for social
phobia. In a study conducted on a Turkish sample of 87 individuals with DSM-III-R
social phobia, 78.2% were men (Gökalp et al., 2001).
Cross-national differences have also been reported in the prevalence of DSM-III
or DSM-IV social phobia. In the Cross-National Epidemiological Surveys there was a
fourfold increase of social phobia in English-speaking Western countries compared to
East Asian countries like Taiwan and South Korea (see Chapman, Mannuzza, & Fyer,
1995). The authors question whether this reflects real differences in the rates of social
phobia across cultures. They note that the interview questions may have lacked cultural
relevance outside Western countries. Also there are conditions analogous to social phobia that are specific to certain Asian countries that were not included in the survey such
as “taijin kyofu-sho” (TKS) in Japan, which is a persistent and irrational fear of causing offense, embarrassment, or harm to others because of some personal inadequacy or
shortcoming (Chapman et al., 1995).
Even within Western countries where rates of social phobia may be quite similar,
the clinical presentation of the disorder can be affected by cultural factors. For example,
a study that compared social phobia in American, Swedish, and Australian samples
found that the Swedish sample was significantly more fearful of eating/drinking in public, writing in public, meetings, and speaking to authority figures (Heimberg, Makris,
Juster, Öst, & Rapee, 1997). Thus social phobia may be found in most countries around
the world but the social concerns, symptom presentation, and even threshold for disorder may vary across cultures (Hofmann & Barlow, 2002; Rapee & Spence, 2004). As
well, the mediating variables for social anxiety may differ between cultures. For example, shame has a stronger mediating role in social anxiety for Chinese than American
samples (Zhong et al., 2008).
Age of Onset and Course
Social phobia typically begins in early to mid-adolescence which gives it a later onset
than specific phobias but an earlier onset than panic disorder (Öst, 1987b; Rapee,
1995a). In the NCS-R, 13 years old was the median age of onset for social phobia which
was substantially younger than the onset age for panic disorder, GAD, PTSD, and OCD
(Kessler, Berglund, et al., 2005). In fact many individuals with social phobia report a
lifelong struggle, with 50–80% reporting an onset of the disorder in childhood (Otto et
al., 2001; Stemberger, Turner, Beidel, & Calhoun, 1995). Early onset is associated with
a more chronic and severe course of the disorder (Beidel & Turner, 2007).
It is commonly believed that untreated social phobia takes a chronic and unremitting course (Beidel & Turner, 2007; Hofmann & Barlow, 2002; Rapee, 1995b). This
appears to be supported by a number of longitudinal studies in which the majority of
Social Phobia
343
individuals with social phobia report a chronic course that can last for years, if not
decades (Chartier, Hazen, & Stein, 1998; Keller, 2003; see Vriends et al., 2007b, for
contrary findings). As with other disorders, it is likely that a greater preponderance of
those with the more chronic form of social phobia will be represented among treatment
seekers.
A number of variables predict chronicity in social phobia. Presence of a comorbid
personality disorder, especially APD, is associated with a lower probability of remission
(Massion et al., 2002), and the generalized subtype of social phobia is characterized by
greater chronicity. Consistent with other anxiety disorders, greater symptom severity
and psychopathology as well as increased functional impairment is associated with a
more enduring and stable course of social phobia (e.g., Chartier et al., 1998; Vriends et
al., 2007b).
Detrimental Effects of Social Phobia
Social phobia is associated with lower educational attainment, lost work productivity,
lack of career advancement, higher rates of financial dependency, and severe impairment in social functioning (e.g., Keller, 2003; Schneier, Johnson, Hornig, Liebowitz,
& Weissman 1992; Simon et al., 2002; Turner, Beidel, Dancu, & Keys, 1986; Zhang,
Ross, & Davidson, 2004). In the NCS individuals with social phobia reported significantly more role impairment than those with agoraphobia but fewer days absent from
work (Magee, Eaton, Wittchen, McGonagle, & Kessler, 1996). As with the other anxiety disorders social phobia with comorbid anxiety (e.g., panic, GAD) or depression has
greater functional impairment (Magee et al., 1996).
Individuals with social phobia also judge their quality of life to be significantly
poorer than nonclinical individuals (Safren, Heimberg, Brown, & Holle, 1996–1997).
A meta-analysis of quality of life in the anxiety disorders revealed that social phobia had
similar negative effects on social, work, and family/home as panic disorder and OCD
(Olatunji et al., 2007). In sum, social phobia is a serious mental disorder that can have
enduring negative effects on life satisfaction and daily living.
Treatment Delay and Service Utilization
Despite many negative effects of the disorder, individuals with social phobia have some
of the lowest rates of treatment utilization of the anxiety disorders. The vast majority of
individuals with social phobia never seek treatment for their condition. In the NCS individuals with social phobia had lower rates of seeking professional help than those with
simple phobia or agoraphobia (Magee et al., 1996). Moreover, only 24.7% of individuals who met DSM-IV criteria for social phobia in the NCS-R made at least one visit to
a mental health specialist in a 12-month period (Wang, Lane, et al., 2005). In the same
study, the median duration of delay in first treatment contact was 16 years for social
phobia, a length of delay that was substantially longer than those for panic disorder,
GAD, PTSD, or major depression (Wang, Berglund, et al., 2005). They also make fewer
general medical visits than individuals with panic disorder (Deacon et al., 2008). In sum
individuals with social phobia are less likely to seek treatment and the minority who
do eventually make an initial contact only after many years with the disorder. Furthermore, social phobia tends to go undetected by physicians and other health professionals,
344
TREATMENT OF SPECIFIC ANXIETY DISORDERS
thereby compounding the problem of low service utilization (Wagner, Silvoe, Marnane,
& Rouen, 2006). In fact analysis of the NCS-R data set also confirms that the disorder
is undertreated, with some evidence that those who have the greatest need for treatment
are least likely to receive it (Ruscio, Brown, et al., 2007).
Clinician Guideline 9.4
Social phobia is the most prevalent of the anxiety disorders, affecting slightly more women
than men, with cultural differences in rate and clinical presentation. The disorder commonly arises in late childhood or adolescence and takes a chronic and unremitting course
that results in significant decrement in social and occupational functioning. Despite these
negative effects, individuals typically delay seeking treatment.
Comorbidity
Social phobia can be difficult to distinguish from other anxiety disorders because social
anxiety is a